CN101459718A - Rubbish voice filtering method based on mobile communication network and system thereof - Google Patents

Abstract

A kind of garbage voice filtering method based on mobile communication network, comprising the steps of: setting the garbage voice reputation threshold value of the mobile terminal; downloading the garbage voice user identifier lower than the garbage voice reputation value in the download server, and saving in the mobile The spam voice user identification information base of the terminal; the calling party identifier of the call request is obtained; the calling party identifier is searched in the spam voice user identification information base, and if the calling user identifier is found, the mobile phone The terminal rejects the call request; otherwise, it is further judged whether it is necessary to query the server for the reputation value of the calling user in real time, and if the reputation value of the calling user is lower than the preset voice spam reputation threshold, the mobile terminal refuses The call request. By analyzing the characteristics of mobile spam voice and the deficiency of existing spam filtering technology, combining the characteristics of mobile communication system to protect mobile communication terminal users from the harassment of spam voice calls.

Description

A mobile communication network-based spam voice filtering method and system thereof

technical field

The invention relates to the field of communication, in particular to a method and system for identifying and filtering junk speech based on a mobile communication network.

Background technique

In recent years, with the rapid development of mobile communications, the penetration rate of mobile phones has repeatedly climbed to new heights, and is gradually moving towards the direction of personal communications. While fully enjoying the convenience provided by mobile communication, people also encounter a series of troubles constantly. A major annoyance is various forms of spam. In particular, when the mobile communication network is connected to the Internet which lacks sufficient supervision, mobile phone users will face more spam voice harassment.

Voice spam refers to voice calls that the called user does not want to receive, and its content includes illegal advertisements, false winning information, and call traps like "ringing" the phone, etc. Because the spam voice is generated through the normal call setup way, it is very difficult for the mobile communication system to filter it. Due to the real-time nature of the voice communication service itself, spam voices are also real-time. When the called party finds that a call is spam voice, the called user has been harassed. Thus, voice spam has a greater disruptive effect on users than spam.

At present, the technologies for filtering spam information such as spam and voice spam can be mainly divided into the following categories:

(1) List-based filtering technology

This type of technology mainly allows end users to establish their own blacklists and whitelists, and the terminals filter or pass them based on the blacklists and whitelists. This technique has a limited effect on preventing spam voice intrusion. The reason is that this technology only relies on the personal experience of the called user, and cannot fundamentally suppress the scope of influence of spam voices.

(2) Content-based filtering technology

This technology is currently mainly used in the field of spam filtering. The terminal uses statistical methods to classify mail according to the content of the mail. This technology divides the email content into words and phrases, and uses Bayesian rule to learn the probability characteristics of each keyword and phrase, so as to judge and filter whether the email is spam. However, since the spam speech is real-time and the speech content cannot be analyzed in advance, this technology is not suitable for the recognition and filtering of spam speech.

(3) Report-based filtering technology

This technology is based on the user's report on spam voice calls, and the report result is shared among all users, so as to realize the filtering of spam information. This kind of technology can only play a normal role under ideal conditions. If the user's report is unfair, not only will it not be able to filter spam voice, but it will also hinder normal voice calls.

(4) Filtering technology based on social network

This technology judges whether a call is a spam voice call by counting the social relationship network of the calling user and the called user, and calculating the social network aggregation coefficient of each user. This technology will mistakenly regard the user terminal of the nature of the service facing the public as a spammer.

According to the above analysis, the existing spam filtering technology is not completely suitable for spam voice filtering in mobile communication networks. Therefore, it is necessary to propose a reliable technology that can effectively filter mobile spam voice in order to improve the utilization rate of mobile network resources. and service quality.

Contents of the invention

The present invention analyzes the characteristics of mobile spam and the deficiencies of the existing spam filtering technology. The present invention combines the characteristics of the mobile communication system, and its purpose is to provide a reputation-based spam filtering method in the mobile communication network to protect mobile communication. End users are protected from spam voice calls.

The object of the present invention is also to provide a spam voice filtering system based on reputation, through which the disturbing voice is filtered.

The object of the present invention is also to provide a spam voice reputation evaluation method, a spam voice reputation evaluation method, and a spam voice reputation sharing method, so as to protect mobile communication terminal users from being harassed by spam voice calls.

The purpose of the present invention and the solution to its technical problems are achieved by adopting the following technical solutions.

A method for filtering spam voice based on mobile communication network, the method comprises the steps of: setting the spam voice reputation threshold value of mobile terminal; downloading the spam voice user identifier lower than the spam voice reputation value in the download server, and saving The spam voice user identification information base in the mobile terminal; the calling party identifier of the call request is obtained; the calling party identifier is searched in the spam voice user identification information base, if the calling party identification is found Then the mobile terminal rejects the call request; otherwise, it is further judged whether it needs to inquire about the calling user reputation value from the server in real time, if the calling user reputation value is lower than the set spam voice reputation threshold value, then mobile The terminal rejects the call request.

The method further includes generating a reputation evaluation at the mobile terminal according to the call model data and call model parameters of the call request.

The method further includes re-evaluating the reputation value of the calling user at the server side according to the reputation evaluation and the historical reputation evaluation of the calling user.

The method further includes the mobile user querying the reputation value of the calling user identifier through the server according to the set real-time query.

The method further includes: when querying the server for the credit value of the calling user in real time, starting a timer.

Wherein the white list saves user identifiers that have made calls with the mobile user.

Wherein, the reputation value of the calling user is stored on the server side, and the calling user is not identified as a voice spam user.

The method further includes the mobile terminal requesting the server to register the identity information.

Wherein the mobile terminal generates the reputation evaluation, further comprising: counting the call model parameters of the mobile user, including the average incoming call duration and outgoing call frequency; if the reputation evaluation generation method is set to automatic mode, then generate reputation according to the following formula evaluate:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}\right.$

为该主叫用户呼叫该移动用户的平均呼叫时长，单位为秒；T_icl为该移动用户接受的所有入呼叫的平均呼叫时长，单位为秒；f_oc为该移动用户呼叫该主叫用户的呼叫频率，单位为次/天；F_oc为该移动用户呼叫任意用户的总的平均呼叫频率，单位为次/天；α和β为权重参数。In the formula, R(T), R(UT) and R(Unknown) constitute a complete reputation evaluation, among which: R(T) represents the credibility of the mobile user, R(UT) represents the untrustworthiness, R(Unknown) represents uncertain;

The average call duration of calling the mobile user for the calling user, in seconds; T _icl is the average call duration of all incoming calls accepted by the mobile user, in seconds; f _oc is the call duration of the mobile user calling the calling user Call frequency, the unit is times/day; F _oc is the total average call frequency of the mobile user calling any user, the unit is times/day; α and β are weight parameters.

Wherein generating the reputation evaluation at the mobile terminal further includes: generating the reputation evaluation according to the following formula according to the subjective evaluation directly provided by the mobile terminal user:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

为该主叫用户呼叫该移动用户的平均呼叫时长，单位为秒；T_icl为该移动用户接受的所有入呼叫的平均呼叫时长，单位为秒；f_oc为该移动用户呼叫该主叫用户的呼叫频率，单位为次/天；F_oc为该移动用户呼叫任意用户的总的平均呼叫频率，单位为次/天；α和β为权重参数；γ为该移动用户的直接评价，γ的取值在0～1之间。In the formula, R(T), R(UT) and R(Unknown) constitute a complete reputation evaluation, among which: R(T) represents the credibility of the mobile user, R(UT) represents the untrustworthiness, R(Unknown) represents uncertain;

The average call duration of calling the mobile user for the calling user, in seconds; T _icl is the average call duration of all incoming calls accepted by the mobile user, in seconds; f _oc is the call duration of the mobile user calling the calling user Call frequency, the unit is times/day; F _oc is the total average call frequency of the mobile user calling any user, the unit is times/day; α and β are weight parameters; γ is the direct evaluation of the mobile user, and the value of γ is The value is between 0 and 1.

The method further includes the mobile terminal evaluating the reputation of the calling user according to the total call model parameters between the mobile terminal user and any user, including the average incoming call duration and outgoing call frequency.

The call model weight parameters α and β can be freely defined by the mobile terminal user, so that the value of R(T) is between 0 and 1.

Wherein, the mobile terminal user defines the value of γ according to three grades of R(T), R(UT) and R(Unknown).

The reputation evaluation provided by the mobile terminal adopts the following format:

{R(T), R(UT), R(Unknown)}

In the formula, R(T) represents credibility, the larger R(T) indicates that the calling user is less likely to be a voice spam user, R(UT) represents unreliability, and the larger R(UT) indicates that the calling user is more likely to be a voice spam user. It may be a spam voice user, R (Unknown) represents uncertainty, and the larger R (Unknown), the more uncertain the mobile user is about the type of the calling user. The values of R(T), R(UT) and R(Unknown) are all between 0 and 1, and the sum of the three is 1.

Wherein, the evaluation of the reputation value of the calling user at the server side further includes: receiving the reputation evaluation provided by the called mobile user, and starting to carry out the reputation evaluation of the calling user; reputation set, then the server attenuates the reputation evaluation provided by the called user according to the following formula to obtain a new reputation evaluation:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

代表提供信誉评价的被叫用户的历史信誉集合，其保存于服务器的信誉数据库。In the formula, {R'(T), R'(UT), R'(Unknown)} represent a new reputation evaluation set,

Represents the historical reputation collection of the called user who provided the reputation evaluation, which is stored in the reputation database of the server.

Wherein the method further utilizes the D-S evidence combination formula of the evidence theory to combine the new reputation evaluation with the historical reputation set of the calling user.

Wherein the method further includes re-evaluating the reputation value of the calling user according to the new reputation set according to the following formula;

$\mathrm{<span\; class="notranslate">\; rep</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span>}$

In the formula, Rep is the reputation value of the calling user, and {R(T), R(UT), R(Unknown)} is the historical reputation collection of the calling user.

The reputation value Rep of the calling user is between -1 and +1.

The method further includes that if the called user does not have a historical reputation set, the server uses a default reputation set to attenuate the received reputation evaluation.

The historical reputation set of the calling user is the result of merging all historical reputation evaluations. When a new reputation evaluation is received, it will be re-evaluated and the historical reputation set will be updated.

The purpose of the present invention and the solution to its technical problems can also be further realized by adopting the following technical measures. A reputation evaluation method, the method comprises the steps of: counting the call model parameters of the mobile user, including average incoming call duration and outgoing call frequency; if the reputation evaluation generation method is set to automatic mode, then generate the reputation evaluation according to the following formula:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}\right.$

为该主叫用户呼叫该移动用户的平均呼叫时长，单位为秒；T_icl为该移动用户接受的所有入呼叫的平均呼叫时长，单位为秒；f_oc为该移动用户呼叫该主叫用户的呼叫频率，单位为次/天；F_oc为该移动用户呼叫任意用户的总的平均呼叫频率，单位为次/天；α和β为权重参数。In the formula, R(T), R(UT) and R(Unknown) constitute a complete reputation evaluation, R(T) represents the credibility of mobile users, R(UT) represents unreliability, and R(Unknown) represents uncertainty ;

The average call duration of calling the mobile user for the calling user, in seconds; T _icl is the average call duration of all incoming calls accepted by the mobile user, in seconds; f _oc is the call duration of the mobile user calling the calling user Call frequency, the unit is times/day; F _oc is the total average call frequency of the mobile user calling any user, the unit is times/day; α and β are weight parameters.

Wherein the method also includes the following steps: the mobile terminal generates reputation evaluation according to the following formula according to the evaluation of the user and the call model parameters:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

为该主叫用户呼叫该移动用户的平均呼叫时长，单位为秒；T_icl为该移动用户接受的所有入呼叫的平均呼叫时长，单位为秒；f_oc为该移动用户呼叫该主叫用户的呼叫频率，单位为次/天；F_oc为该移动用户呼叫任意用户的总的平均呼叫频率，单位为次/天；α和β为权重参数，γ为该移动用户的直接评价，γ的取值在0～1之间。In the formula, R(T), R(UT) and R(Unknown) constitute a complete reputation evaluation, R(T) represents the credibility of mobile users, R(UT) represents unreliability, and R(Unknown) represents uncertainty ;

The average call duration of calling the mobile user for the calling user, in seconds; T _icl is the average call duration of all incoming calls accepted by the mobile user, in seconds; f _oc is the call duration of the calling user for the mobile user Call frequency, the unit is times/day; F _oc is the total average call frequency of the mobile user calling any user, the unit is times/day; α and β are weight parameters, γ is the direct evaluation of the mobile user, and the value of γ is The value is between 0 and 1.

The purpose of the present invention and its technical problems can also be further realized by adopting the following technical measures. A kind of reputation assessment method, this method comprises the following steps: receive the reputation assessment of mobile terminal, begin to carry out reputation assessment to calling user; The reputation evaluation provided by the mobile terminal user is attenuated to obtain a new reputation evaluation:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

代表信誉评价提供用户的历史信誉集合；In the formula, {R'(T), R'(UT), R'(Unknown)} represent a new reputation evaluation set,

Provide the user's historical reputation collection on behalf of the reputation evaluation;

The purpose of the present invention and its technical problems can also be further realized by adopting the following technical measures. A reputation sharing method downloads the spam voice user identifiers below the reputation threshold from a server to a mobile terminal to achieve reputation sharing.

Wherein the mobile terminal sets different reputation thresholds according to different requirements for spam voice filtering, and then downloads different numbers of spam voice user identifiers from the server.

The purpose of the present invention and its technical problems can also be further realized by adopting the following technical measures. A spam voice filtering system based on a mobile communication network, the system includes: a mobile terminal and a server, and the mobile terminal includes: a human-computer interaction unit for setting the spam voice reputation value of the mobile terminal; a spam voice user identification information Library, used to save the spam voice user identifier; a white list, used to save the trustworthy user identifier that the mobile user once called; a mobile state monitoring unit, used to monitor incoming call and outgoing call parameters; a spam voice Filter unit, filter the caller identifier in the call request according to the spam voice user identifier in the spam voice user identifier information base; a spam voice user identifier download unit, used to download the latest garbage voice from the server User identifier information; a call parameter statistical unit, according to the mobile state monitoring unit monitoring call parameters to count the average duration and call frequency of all incoming and outgoing calls of the mobile terminal; a garbage voice reputation evaluation unit, according to the call The parameters counted by the parameter statistical unit are used to generate reputation evaluation for the calling user of the incoming call; a user identity unit is used to complete the identity authentication of the mobile terminal user on the server side; and a protocol communication unit is used to provide mobile Protocol communication function between terminal and server. The server end includes: a reputation database for storing mobile terminal user identity information and reputation data; a reputation evaluation unit for performing reputation evaluation on the calling user according to the new reputation evaluation and historical reputation set sent by the called mobile terminal. evaluation; a reputation query unit, used for querying the reputation database, and fulfilling the reputation query request put forward by the mobile terminal; and a communication unit, used for protocol communication between the server and the mobile terminal.

Wherein the reputation evaluation unit further includes: a reputation evaluation attenuation module, used to reasonably attenuate the reputation evaluation received by the server; a reputation evaluation merging module, used to combine the reputation evaluation and historical reputation received by the server The combination calculation of the sets obtains the latest reputation set; a reputation quantification module calculates the reputation value for identifying spam voice users according to the reputation set.

Wherein said credit query unit further comprises: a real-time query module, used for responding to the real-time credit query request that mobile terminal proposes, and this request only inquires the reputation of the single user of designation; Spam user identifier download request.

Wherein said spam voice user identification information base further includes: an information storage management module, used for storing and managing user spam voice user identifier data, and also used for indexing of spam voice user identifiers; an information access module, used for An interface for accessing the information storage management unit is provided to the spam voice filtering unit and the spam voice reputation evaluation unit, and the interface includes query, insertion and deletion.

The protocol communication unit further includes: a communication protocol module for decoding received data and encoding for sent data; a message communication module for sending and receiving data between the mobile terminal and the server.

By virtue of the above technical solutions, the present invention has at least the following advantages:

1. The present invention utilizes the spam voice user identifier stored in the mobile terminal and the method of inquiring the reputation of the calling user in real time, and has the advantage of real-time filtering of spam voice.

2. The server of the present invention can be used as an integral part of the communication system in the mobile communication network, and can also be independent of the existing mobile communication system. The deployment of the server has the advantage of not changing the existing mobile communication network topology and the existing mobile communication system.

3. The reputation evaluation method of the present invention utilizes the reputation evaluation provided by the user to attenuate the reputation evaluation provided by the user, so that the reputation evaluation has the advantages of fairness and justice.

4. The present invention integrates the direct evaluations of numerous mobile users, has the advantage of quickly and accurately identifying garbage voices, and greatly reduces the influence range of garbage voices.

5. The present invention has the advantage of personalized spam voice filtering by setting different reputation thresholds for mobile users, and mobile users can according to different situations.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable , the following preferred embodiments are specifically cited below, and are described in detail as follows in conjunction with the accompanying drawings.

Description of drawings

Fig. 1 is a schematic diagram of the network structure of the mobile spam voice filtering system of the present invention.

Fig. 2 is a structural schematic diagram of the mobile spam voice filtering system of the present invention.

Fig. 3 is a flow diagram of the method for filtering mobile spam voices of the present invention.

Fig. 4 is a schematic flow chart of mobile user identity information registration in the present invention.

Fig. 5 is a schematic flow chart of spam voice reputation evaluation in the present invention.

Fig. 6 is a schematic flow chart of spam voice reputation evaluation in the present invention.

Fig. 7 is a schematic flow chart of downloading the spam voice user identifier in the present invention.

Detailed ways

In order to further explain the technical means and effects of the present invention to achieve the intended purpose of the invention, the specific implementation, structure, features and effects of the wireless positioning method proposed according to the present invention will be described below in conjunction with the accompanying drawings and preferred embodiments. Details are as follows.

The core idea of the present invention is to implement the spam voice filtering based on the reputation evaluation of the mobile terminal user without changing the topological structure of the mobile communication network and the internal structure of the mobile communication equipment. Embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

Fig. 1 is a network structure diagram of the mobile spam voice filtering system of the present invention. The server 11 shown in Fig. 2 does not belong to the internal equipment of the mobile communication system, but accesses the mobile communication network through IP and SMS. Multiple servers 11 may be included in a set of mobile garbage voice filtering system, and form a distributed system. The client in the mobile spam filtering system, that is, the mobile terminal 12 accesses the mobile communication network in a wireless manner, and communicates with the server 11 .

Fig. 2 is the block diagram of the mobile spam voice filtering system of the present invention, has described in detail on the basis of Fig. 1 the internal unit structure of the server 11 and the mobile terminal 12 forming the mobile spam voice filtering system, and the relationship between the internal units.

Mobile terminal 12 comprises: man-machine interaction unit 121, spam voice filtering unit 122, spam voice reputation evaluation unit 123, spam voice user identification download unit 124, spam voice user identification information storehouse 125, white list 126, call parameter statistics unit 127, User identity unit 128 , mobile terminal status monitoring unit 129 and protocol communication unit 1210 .

The human-computer interaction unit 121 is used to display the output information of the spam voice filtering unit 122 and the spam voice reputation evaluation unit 123, and is also used to display the information input by the spam voice reputation evaluation unit 123 received by the user, and to set the spam voice reputation value.

Spam voice filtering unit 122 is used to obtain the caller identifier in the call request, and inquires the caller identifier in the spam voice user identifier storehouse 125 of the mobile terminal, if the caller identifier is found, then filter the call request . If do not find this caller identifier in spam voice user identification storehouse 125, then spam voice filtering unit 122 determines whether to send to the server by protocol communication unit 1210 by the real-time reputation query item that white list 126 and human-computer interaction unit 121 set 11. Inquire about the reputation of the calling user in real time.

The spam voice reputation evaluation unit 123 is used to generate reputation evaluations for incoming callers according to the results of the call parameter statistics unit 127. The reputation evaluation evaluation generation methods include automatic and manual methods. The automatic and manual calculation methods for specific evaluations are described below . The evaluation criteria provided by the spam voice reputation evaluation unit 123 to the user are divided into three levels: credible, general, and unreliable. The reputation evaluation generated by the spam voice reputation evaluation unit 123 includes the above three levels.

The spam user identifier downloading unit 124 queries and downloads the latest spam user identifier information from the reputation query unit 113 of the server 11 through the protocol communication unit 1210 .

The spam voice user identification information base 125 saves part of the spam voice user identifier and the spam voice user identifier downloaded from the spam voice identification download unit 124, wherein the reputation of the calling user identified by the user identifier stored is lower than that of the mobile terminal 12. A predetermined reputation threshold, which is an arbitrary value between -1 and +1 set by the mobile terminal user through the human-computer interaction unit 121 .

Wherein the spam voice user identification library 125 further includes an information storage management module and an information access module (not shown). The information storage management module saves and manages user spam user identifier data, and can also be used for indexing spam user identifiers. The spam voice user identifier data stored in the information storage management module includes user identifiers. The information access module is used to provide the spam voice filtering unit 122 and the spam voice reputation evaluation unit 123 with an interface for accessing the information storage management unit, and the interface includes query, insertion and deletion processing functions.

The call parameter statistics unit 127 makes statistics on the call model parameters of the incoming call and outgoing call of the mobile terminal, specifically statistics: incoming call duration, outgoing call duration, incoming call frequency and outgoing call frequency. These four call parameters are respectively used for the total call model of the mobile terminal and for each specific calling user and called user to perform statistics on the call model parameters. The statistical results of the call parameter statistics unit 127 are provided to the spam reputation evaluation unit 123 to complete the calculation of the reputation evaluation of the called user, and the specific calculation method is described below.

The white list 126 saves the identifiers of users who have been evaluated by the spam voice reputation evaluation unit 123 as trustworthy with which the mobile user has ever called.

The user identity unit 128 generates the identity information of the mobile terminal user, and cooperates with the server to complete the verification of the uniqueness of the user identity. The user identity information is calculated by the mobile user identifier and the mobile terminal equipment identifier through a message digest encryption algorithm (wherein the encryption algorithm can be such as the existing MD5 algorithm or SHA-1 algorithm, etc.).

The mobile terminal state monitoring unit 129 monitors the state of the mobile terminal 12. When the mobile terminal 12 received a call request, the mobile terminal state monitoring unit 129 was responsible for obtaining the calling party identifier, and passed the calling party identifier to the spam voice filtering Unit 122 processes. When the mobile terminal state monitoring unit 129 detects that the mobile terminal 12 terminates an incoming call, the event is sent to the spam voice reputation evaluation unit 123 for processing through the spam voice filtering unit 122, and the calling user is evaluated for reputation, the specific evaluation method Described below. When the mobile terminal state monitoring unit 129 detects that the mobile terminal 12 initiates a call, the called user identifier is extracted from the mobile terminal 12 and the spam voice reputation evaluation unit 123 stores the called user identifier in the white list 126 .

The protocol communication unit 1210 is used to provide protocol communication between the mobile terminal 12 and the server 11, and is also used for real-time query of the reputation of the calling user and download of the spam user identifier.

The protocol communication unit 1210 further includes a communication protocol module and a message communication module (not shown). The communication protocol module is used to decode received data and encode sent data. The message communication module is used for sending and receiving data between the mobile terminal and the server.

The server 11 shown in FIG. 2 includes a reputation evaluation unit 112 , a reputation query unit 113 , a communication unit 114 , a reputation database 115 and a user identity management unit 111 .

The reputation evaluation unit 112 receives the latest reputation evaluation from the spam reputation evaluation unit 123 via the protocol communication unit 1210 of the mobile terminal 12, and combines the latest reputation evaluation with the historical reputation set.

The reputation evaluation unit 112 further includes a reputation evaluation attenuation module, a reputation evaluation merging module and a reputation value quantization module (not shown). The reputation evaluation attenuation module reasonably attenuates the reputation evaluation received by the server 11, and its specific calculation method is described below, so as to avoid the impact of malicious evaluation on the objectivity of reputation evaluation. The reputation evaluation merging module is used to combine and calculate the reputation evaluation received by the server from the spam voice reputation evaluation unit 123 and the historical reputation set to obtain the latest reputation set. The reputation value quantification module calculates the reputation value used to identify spam voice users according to the newly merged reputation set, and the reputation value is between -1 and +1.

The reputation query unit 113 completes the reputation query request made by the mobile terminal 12 .

The reputation query unit 113 further includes a real-time query module and a non-real-time download module (not shown). The real-time query module is used to respond to the real-time reputation query request made by the mobile terminal 12, and the request only queries the reputation of a specified single user. The non-real-time downloading module is used for completing the request for downloading the spam voice user identifier proposed by the mobile terminal 12 .

The communication unit 114 is used to implement protocol communication between the server 11 and the mobile terminal 12, and is also used to decode received data and encode sent data.

The reputation database 115 stores user reputation data, which includes user identifiers, user reputation values, and user reputation sets. The user reputation set further includes three parts: trusted, unknown and untrusted.

The user identity management unit 111 receives the registration request of the mobile terminal 12 and manages the user identity.

Fig. 3 is the flow process of filtering garbage speech method in the mobile communication network provided by the present invention, and this flow process specifically comprises the following steps:

Step 301, the mobile terminal 12 initializes a part of the spam user identifier library stored locally, and sets the spam reputation threshold through the human-computer interaction unit 121;

Step 302, the mobile terminal 12 downloads the spam voice user identifier lower than the reputation threshold set in step 301 from the server 11, and stores it in the local spam voice user identifier storehouse;

Step 303, the state monitoring unit 129 of the mobile terminal 12 monitors whether a call request arrives, if a new call arrives, execute step 304, otherwise continue to monitor the call request;

Step 304, the mobile terminal 12 extracts the calling party identifier from the received call request, and inquires in the part of the spam voice user identification information storehouse 125 stored locally;

Step 305, if the mobile terminal 12 finds the calling party identifier in the part of the spam voice user identifier storehouse 125 stored locally, execute step 315, otherwise execute step 306;

Step 306, the mobile terminal 12 determines whether real-time credit query needs to be performed according to whether the mobile user sets the option of real-time query reputation and whether the mobile terminal user has actively communicated with the calling user, if necessary, execute step 311, otherwise execute Step 307;

Step 307, the mobile terminal 12 prompts the user that the calling user of the call request is not a voice spam user, and prompts the user to accept the call request;

Step 308, monitor whether the call has ended, if yes, execute step 309, otherwise continue monitoring;

Step 309, the mobile terminal 12 generates the reputation evaluation of the calling user according to the spam voice reputation evaluation method (the specific evaluation method is described below), and sends it to the server 11;

Step 310, the server 11 receives the reputation evaluation of the mobile terminal 12, re-evaluates the calling user, and saves the new reputation value;

Step 311, the mobile terminal 12 sends a request to the server 22 for real-time query reputation, which includes the user's identity information and the calling user identifier, and starts timer T1 to monitor whether the query result is received on time, so as to ensure that the garbage Real-time performance of voice filtering service;

Step 312, the server 11 searches the credit value of the calling user in the reputation database 115 according to the user identity information and the calling user identifier in the query request, and sends it to the mobile terminal 12;

Step 313, if the timer T1 started by the mobile terminal 12 expires, execute step 307, otherwise execute step 314;

Step 314, the mobile terminal 12 extracts the reputation value of the calling user from the received reputation query result, and compares it with the reputation threshold value set by the mobile user, if it is higher than the preset reputation threshold value, execute step 307, Otherwise, execute step 315 .

In step 315, the mobile terminal 12 confirms that the calling user is a voice spam user, rejects the call request, and ends the process.

Fig. 4 is the flowchart of mobile user identity information registration of the present invention, and this flow process specifically comprises the following steps:

Step 401, the mobile terminal 12 extracts the identifier of the local mobile user and the identifier of the local mobile terminal equipment;

Step 402, setting an identity password for the mobile user;

Step 403, the mobile terminal 12 applies a message digest algorithm (such as the existing MD5 algorithm or SHA-1 algorithm, etc.) to calculate the identity information digest based on the user's identifier, the device identifier of the mobile terminal and the identity password, and the mobile terminal 12 Subsequent message exchange with the reputation evaluation unit 112 and the reputation query unit 113 of the server 11 only provides the identity information summary, which not only ensures the uniqueness of the user identity in the server 11, but also achieves the purpose of protecting the privacy of the mobile user;

Step 404, the mobile terminal 12 sends the summary of the identity information of the mobile terminal calculated in step 403 to the server 11, requesting registration;

Step 405, if the server 11 receives the registration request of the mobile terminal, the user identity management unit 111 executes step 406, otherwise continue to wait for the registration request;

Step 406, the user identity management unit 111 of the server 11 extracts the user identity information summary from the received registration request, and judges its uniqueness;

Step 407, if the identity information of the mobile user is unique, execute step 408, otherwise execute step 409;

Step 408, the user identity management unit 111 of the server 11 stores the summary of the mobile user identity information into the reputation database 115;

Step 409, the server 11 sends the registration result to the mobile terminal 12 requesting registration, if the identity information of the mobile user is unique, it returns the registration success, otherwise it sends the registration failure.

Fig. 5 is the flow chart of rubbish voice reputation evaluation among the present invention, and concrete steps comprise:

Step 501, setting the reputation evaluation generation method in the mobile terminal 12, the specific generation method includes: automatic method and manual method, the automatic method is generated by the automatic calculation of the mobile terminal according to the call model parameters, and the manual method is subjectively evaluated by the called mobile user, Combined with the calculation of the call model parameters;

Step 502, the mobile terminal 12 monitors whether the mobile user enters a call state, if the call starts, execute step 503, otherwise continue to monitor the state of the mobile terminal device;

Step 503, the mobile terminal 12 judges whether the call is an incoming call request, if yes, execute step 505, otherwise execute step 504;

Step 504, the mobile terminal 12 records the called mobile subscriber identifier of the call into the white list 126, and records the call start time;

Step 505, the mobile terminal 12 monitors whether the call ends, if the call ends, execute step 506, otherwise continue monitoring;

Step 506, the mobile terminal 12 records the call end time, and counts the duration of the call;

Step 507, the mobile terminal 12 judges the reputation evaluation generation method set by the called mobile user, if it is an automatic method, execute step 508, otherwise execute step 509;

Wherein the reputation evaluation mode provided by the mobile terminal 12 adopts the following format:

{R(T), R(UT), R(Unknown)}

In the formula, R(T) represents credibility, the larger R(T) indicates that the calling user is less likely to be a voice spam user, R(UT) represents unreliability, and the larger R(UT) indicates that the calling user is more likely to be a voice spam user. It may be a spam voice user, R (Unknown) represents uncertainty, and the larger the R (Unknown), the less certain the type of the called mobile user is to the calling user. The values of R(T), R(UT) and R(Unknown) are all between 0 and 1, and the sum of the three is 1;

Step 508, the mobile terminal 12 automatically generates reputation evaluation according to the call model parameters, and executes step 512;

Provide below this automatic mode in step 508 and carry out reputation evaluation formula:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}\right.$

为该主叫用户呼叫该被叫移动用户的平均呼叫时长，单位为秒；T_icl为该被叫移动用户接受的所有入呼叫的平均呼叫时长，单位为秒；f_oc为被叫该移动用户呼叫该主叫用户的呼叫频率，单位为次/天；F_oc为该被叫移动用户呼叫任意用户的总的平均呼叫频率，单位为次/天；α和β为权重参数可以由被叫移动用户通过人机交互单元121自由定义，使R(T)取值在0～1之间。其中

、T_icl、f_oc和F_oc由呼叫参数统计单元127统计、计算而得。In the formula, R(T), R(UT) and R(Unknown) constitute a complete reputation evaluation, R(T) represents credibility, R(UT) represents unreliability, and R(Unknown) represents uncertainty;

The average call duration of calling the called mobile user for the calling user, in seconds; T _icl is the average call duration of all incoming calls accepted by the called mobile user, in seconds; f _oc is the called mobile user The call frequency of calling the calling user, the unit is times/day; F _oc is the total average call frequency of the called mobile user calling any user, the unit is times/day; α and β are weight parameters that can be determined by the called mobile The user freely defines through the human-computer interaction unit 121, so that the value of R(T) is between 0 and 1. in

, T _icl , f _oc and F _oc are calculated and obtained by the call parameter statistics unit 127 .

Step 509, the mobile terminal 12 generates a reputation evaluation according to the called mobile user's subjective evaluation and call model parameters, which generates a reputation evaluation according to the following formula:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\frac{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; t</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; icl</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; oc</span>}}}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

、T_icl、f_oc、F_oc、α和β与步骤508的信誉评价计算公式中的符号具有相同的含义；γ为该移动用户的直接评价，γ的取值在0～1之间。移动用户的直接评价包括三个等级：可信、不可信和不确定。移动用户选择不同的评价等级，γ取不同的值。where R(T), R(UT), R(Unknown),

, T _icl , f _oc , F _oc , α and β have the same meanings as the symbols in the reputation evaluation calculation formula in step 508; γ is the direct evaluation of the mobile user, and the value of γ is between 0 and 1. The direct evaluation of mobile users includes three grades: trustworthy, untrustworthy and uncertain. Mobile users choose different evaluation grades, and γ takes different values.

Step 510, whether the mobile user inputs the direct evaluation of "unreliable" that the calling user is a spam voice user through the human-computer interaction unit 121, if yes, execute step 511, otherwise execute step 512;

Step 511, the mobile terminal 12 stores the calling party identifier in the local spam voice user identification library 125, and ends this process;

Step 512, the mobile terminal 12 sends a reputation evaluation message to the server 11, wherein the message includes the user identity information summary of the called mobile user, the calling user identifier and the reputation evaluation, and the flow ends.

Fig. 6 is the flow chart of spam voice reputation evaluation among the present invention, and concrete steps comprise:

Step 601, the server 11 monitors various input events, which is specifically completed by the communication unit 114 of the server 11;

Step 602, the communication unit 114 judges whether the server 11 has received the reputation evaluation from the mobile terminal 12, which includes the reputation evaluation {R(T), R(UT), R(Unknown)}, if yes, execute step 603, otherwise continue monitor input events;

Step 603, the reputation evaluation unit 112 of the server 11 extracts the identity information abstract of the called mobile user who sent the reputation evaluation from the received reputation evaluation message, and queries the reputation database to verify whether the called mobile user has passed the registration ;

Step 604, the server 11 verifies the identity information of the called mobile user through the reputation database 115, if the identity information of the called mobile user is verified as legal, execute step 606, otherwise execute step 605;

Step 605, the reputation evaluation unit 112 of the server 11 discards the reputation evaluation, and ends this process;

Step 606, the reputation evaluation unit 112 of the server 11 extracts the reputation evaluation from the received reputation evaluation message, and verifies the legitimacy of the evaluation. The verification method is to judge the credibility, unreliability and uncertainty contained in the reputation evaluation Whether the sum of the three is 1, that is, the sum of the three including reputation evaluation {R(T), R(UT), R(Unknown)} is 1;

Step 607, the reputation evaluation unit 112 of the server 11 judges that if the reputation evaluation is legal, execute step 608, otherwise execute step 605;

Step 608, the reputation evaluation unit 112 of the server 11 searches the credit value in the reputation database 115 according to the identity information abstract of the called mobile user;

In step 609, the credit evaluation unit 112 of the server 11 judges that if the reputation of the called mobile user is found, execute step 610, otherwise execute step 614;

Step 610, the reputation evaluation unit 112 of the server 11 utilizes the reputation obtained through the reputation evaluation attenuation module of the reputation evaluation unit 112 to attenuate the reputation evaluation provided by the called mobile user, so as to avoid malicious evaluation by a small number of users;

The server attenuates the reputation evaluation provided by the called mobile user according to the following formula to obtain a new reputation evaluation

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span>\end{array}\right.$

where {R'(T), R'(UT), R'(Unknown)} represents a new reputation evaluation, Provide user's historical reputation collection on behalf of reputation evaluation, this historical reputation collection is stored in the reputation database 115, if there is no historical reputation collection of the called mobile user in the reputation database, then the server adopts the default reputation collection to provide the called mobile user For example, the default reputation set is: {0.3, 0.3, 0.4}, here is an example but not limited thereto;

Step 611, the reputation evaluation unit 112 of the server 11 extracts the identifier of the calling user of the evaluation object from the received reputation evaluation message, and queries the historical reputation of the calling user in the reputation database 115, and the historical reputation also includes: Reliability, unreliability and uncertainty;

Step 612, the reputation combining module of the reputation evaluation unit 112 in the server 11 merges the reputation evaluation and historical reputation to achieve the purpose of re-evaluating the calling user's reputation;

Wherein the reputation combining module of the server 11 utilizes the D-S evidence combining formula of the evidence theory to combine the above-mentioned calculated new reputation evaluation with the historical reputation set of the calling user, and combine the calculated reputation set according to the following formula for the calling user: Call the user's reputation value to re-evaluate:

$\mathrm{<span\; class="notranslate">\; rep</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; UT</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Unknown</span>}<span\; class="notranslate">\; )</span>}$

In the formula, Rep is the reputation value of the calling user, which is between -1 and +1, and {R(T), R(UT), R(Unknown)} is the historical reputation collection of the calling user. The closer the Rep value is to +1, the higher the reputation of the calling user; on the contrary, the closer the Rep is to -1, the lower the reputation of the calling user;

Step 613, the reputation value quantification module of the server 11 quantifies the reputation value of the evaluation object according to the combined new reputation set, that is, the reputation value quantification module calculates the reputation value for identifying spam users according to the latest merged reputation set, The reputation value is between -1 and +1.

Step 614, the server 11 stores the new reputation calculated in step 613 into the reputation database, and ends this process.

Fig. 7 is the flow chart of downloading the garbage voice user identifier among the present invention, and concrete steps comprise:

Step 701, the mobile terminal 12 monitors the state of the local device, only when the local mobile terminal device is idle, the download process of the spam voice user identifier is started;

Step 702, the mobile terminal state monitoring unit 129 of the mobile terminal 12 judges whether the mobile terminal is in a call state, if not, execute step 703, otherwise continue to monitor the device state;

Step 703, the mobile user inputs the reputation threshold through the man-machine interaction unit 121 of the mobile terminal 12, and sends the spam voice user identifier download request to the server 11 by the spam voice user identifier downloading unit 124. Time and a summary of the mobile user's identity information;

Step 704, the communication unit 114 of server 11 judges whether to receive the spam voice user identifier download request, if receive request, execute step 705, otherwise continue to wait for the download request that mobile terminal 12 sends;

Step 705, the server 11 extracts the mobile user identity information abstract from the received spam voice user identifier download request, and hands over the user identity management unit 111 to query the user's identity information in the reputation database 115;

Step 706, the user identity management unit 111 of the server 11 verifies whether the mobile user identity information is legal, if verified, the credit query unit 113 executes step 708, otherwise executes step 707;

Step 707, the credit query unit 113 of the server 11 discards the spam voice user identifier download request, and ends the process;

Step 708, the reputation query unit 113 of the server 11 extracts the reputation threshold value from the received spam voice user identifier download request, and inquires the reputation value of the spam voice user identifier lower than the reputation threshold value in the reputation database 115;

Step 709, the server 11 sends the searched spam user identifier to the mobile terminal 12 through the communication unit 114;

Step 710, the mobile terminal 12 waits for the query result returned by the server 11, if the query result is received, execute step 711, otherwise execute step 712;

Step 711, the mobile terminal 12 stores the received spam user identifier into the local spam user identification storehouse 125, and ends the process;

Step 712, the mobile terminal 12 resends the spam user identifier download request, and executes step 704.

The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. Anyone familiar with this field Those skilled in the art, without departing from the scope of the technical solution of the present invention, can use the method and technical content disclosed above to make some changes or modifications to equivalent embodiments with equivalent changes, but any content that does not depart from the technical solution of the present invention, Any simple modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims (39)

1, a kind of rubbish voice filtering method based on mobile radio communication is characterized in that, this method comprises the steps:

Set the rubbish voice prestige threshold values of portable terminal;

Have the rubbish voice user identifier that is lower than described rubbish voice credit value in the Download Server, and be stored in the rubbish voice user totem information storehouse of portable terminal;

Obtain the caller ID of call request;

In described rubbish voice user totem information storehouse, search described caller ID, if find described caller ID then portable terminal refuse this call request;

Otherwise further judging whether need be in real time to the described calling subscriber's credit value of described server lookup, if described calling subscriber's credit value is lower than the described rubbish voice prestige threshold values of setting, then portable terminal is refused this call request.

2, method according to claim 1 is characterized in that described method also comprises: when described call request finishes, according to call model data and the call model parameter of described call request, at credit rating of described portable terminal generation.

3, as method as described in the claim 2, it is characterized in that described method also comprises:, again described calling subscriber is carried out the assessment of credit value at server end according to described credit rating and calling subscriber's historical credit rating.

4, method according to claim 1 is characterized in that described method also comprises: the mobile subscriber inquires about the credit value of described caller ID by server according to the real-time inquiry of setting.

5, as method as described in the claim 4, it is characterized in that described method also comprises: when to the described calling subscriber's credit value of described server inquiry in real time, start a timer.

As method as described in the claim 4, it is characterized in that 6, described white list is preserved with described mobile subscriber and carried out the user identifier called out.

7, method according to claim 1 is characterized in that, preserves calling subscriber's credit value at server end, and not identify this calling subscriber be the rubbish voice user.

8, method according to claim 1, it is characterized in that described method also comprises: described portable terminal is registered to the server end requests identity information.

9, as method as described in the claim 2, it is characterized in that, produce described credit rating, further comprise at described portable terminal:

Add up this mobile subscriber's call model parameter, comprise and on average go into call duration and go out calling frequency;

If the credit rating producing method is set at automated manner, then produce credit rating according to following formula,

$\left\{\begin{array}{c}R\left(T\right)=\mathrm{\&alpha;}\frac{{\stackrel{\&OverBar;}{t}}_{\mathrm{icl}}}{{\stackrel{\&OverBar;}{T}}_{\mathrm{icl}}}+\mathrm{\&beta;}\frac{f_{\mathrm{oc}}}{{\stackrel{\&OverBar;}{F}}_{\mathrm{oc}}}\\ R\left(\mathrm{UT}\right)=\frac{1-R\left(T\right)}{2}\\ R\left(\mathrm{Unknown}\right)=\frac{1-R\left(T\right)}{2}\end{array}\right.$

R in the formula (T), R (UT) and R (Unknown) constitute complete credit rating, and wherein: R (T) expression mobile subscriber's confidence level, R (UT) expression can not reliabilities, and R (Unknown) expression is uncertain; For this calling subscriber calls out this mobile subscriber's average call duration, unit is second; T _IclBe all average call durations of going into to call out that this mobile subscriber accepts, unit is second; f _OcFor this mobile subscriber calls out this calling subscriber's calling frequency, unit is time/day; F _OcBe total average call frequency that this mobile subscriber calls out any user, unit is time/day; α and β are weight parameter.

10, as method as described in the claim 2, it is characterized in that, produce described credit rating, further comprise at described portable terminal:

Subjective assessment according to mobile phone users directly provides produces credit rating according to following formula:

$\left\{\begin{array}{c}R\left(T\right)=\mathrm{\&alpha;}\frac{{\stackrel{\&OverBar;}{t}}_{\mathrm{icl}}}{{\stackrel{\&OverBar;}{T}}_{\mathrm{icl}}}+\mathrm{\&beta;}\frac{f_{\mathrm{oc}}}{{\stackrel{\&OverBar;}{F}}_{\mathrm{oc}}}\\ R\left(\mathrm{UT}\right)=[1-R\left(T\right)]\mathrm{\&gamma;}\\ R\left(\mathrm{Unknown}\right)=[1-R\left(T\right)](1-\mathrm{\&gamma;})\end{array}\right.$

R in the formula (T), R (UT) and R (Unknown) constitute complete credit rating, and wherein: R (T) expression mobile subscriber's confidence level, R (UT) expression can not reliabilities, and R (Unknown) expression is uncertain;

For this calling subscriber calls out this mobile subscriber's average call duration, unit is second; T _IclBe all average call durations of going into to call out that this mobile subscriber accepts, unit is second; f _OcFor this mobile subscriber calls out this calling subscriber's calling frequency, unit is time/day; F _OcBe total average call frequency that this mobile subscriber calls out any user, unit is time/day; α and β are weight parameter; γ is this mobile subscriber's direct evaluation, and the value of γ is between 0～1.

11, as claim 9 or 10 described methods, it is characterized in that, comprise that further described portable terminal according to the total call model parameter between this mobile phone users and any user, also comprises and on average goes into call duration and go out calling frequency, finish evaluation calling subscriber's prestige.

12, as claim 9 or 10 described methods, it is characterized in that weight parameter α and β can freely be defined by mobile phone users, make R (T) value between 0～1.

13, method as claimed in claim 10 is characterized in that, mobile phone users comes the value of self-defined γ according to R (T), R (UT) and R (Unknown) Three Estate.

14, as claim 9 or 10 described methods, it is characterized in that the credit rating that portable terminal provides adopts following form:

{R(T)，R(UT)，R(Unknown)}

In the formula, R (T) represents confidence level, the big more expression of R (T) calling subscriber is impossible more to be the rubbish voice user, R (UT) representative can not reliability, the big more expression of R (UT) calling subscriber may be the rubbish voice user more, R (Unknown) represents uncertainty, and the big more expression of R (Unknown) mobile subscriber can not determine more to calling subscriber's type.The value of R (T), R (UT) and R (Unknown) is all between 0～1, and three's sum is 1.

15, method as claimed in claim 3 is characterized in that, again the assessment that described calling subscriber carries out credit value is further comprised at server end:

Receive the credit rating that the callee provides, begin the calling subscriber is carried out the prestige assessment;

If preserved this called subscriber's prestige set in the credit database of server, then the credit rating that this called subscriber provided according to following formula of server is decayed and is obtained new credit rating:

$\left\{\begin{array}{c}R\&prime;\left(T\right)=R\left(T\right)\&times;\tilde{R}\left(T\right)\\ R\&prime;\left(\mathrm{UT}\right)=R\left(\mathrm{UT}\right)\&times;\tilde{R}\left(T\right)\\ R\&prime;\left(\mathrm{Unknown}\right)=1-R\&prime;\left(T\right)-R\&prime;\left(\mathrm{UT}\right)\end{array}\right.$

{ R ' (T), R ' (UT), R ' is (Unknown) } new credit rating set of expression in the formula,

Representative provides the called subscriber's of credit rating historical prestige set, and it is stored in the credit database of server.

16, method as claimed in claim 15 is characterized in that, further utilizes the D-S evidence of evidence theory to merge formula the historical prestige of described new credit rating with this calling subscriber is merged, and obtains new credit rating and gathers.

17, method as claimed in claim 16 is characterized in that, further comprises described new credit rating set is reappraised to this calling subscriber's credit value according to following formula;

$\mathrm{Rep}=\frac{R\left(T\right)}{R\left(T\right)+R\left(\mathrm{Unknown}\right)}-\frac{R\left(\mathrm{UT}\right)}{R\left(\mathrm{UT}\right)+R\left(\mathrm{Unknown}\right)}$

Rep is this calling subscriber's a credit value in the formula, and { R (T), R (UT), R (Unknown) } is this calling subscriber's historical prestige set.

18, method as claimed in claim 17 is characterized in that, calling subscriber's credit value Rep is between-1～+ 1.

19, method as claimed in claim 18 is characterized in that, comprises that further server is decayed to the credit rating of receiving with default prestige set if the called subscriber does not have historical prestige set.

20, method as claimed in claim 19 is characterized in that, this calling subscriber's historical prestige set is the result after being merged by all historical credit ratings, when receiving new credit rating, will reappraise and upgrade historical prestige set.

21, a kind of reputation evaluation method based on mobile radio communication is characterized in that, this method comprises the steps:

Add up this mobile subscriber's call model parameter, comprise and on average go into call duration and go out calling frequency;

If the credit rating producing method is set at automated manner, then produce credit rating according to following formula:

$\left\{\begin{array}{c}R\left(T\right)=\mathrm{\&alpha;}\frac{{\stackrel{\&OverBar;}{t}}_{\mathrm{icl}}}{{\stackrel{\&OverBar;}{T}}_{\mathrm{icl}}}+\mathrm{\&beta;}\frac{f_{\mathrm{oc}}}{{\stackrel{\&OverBar;}{F}}_{\mathrm{oc}}}\\ R\left(\mathrm{UT}\right)=\frac{1-R\left(T\right)}{2}\\ R\left(\mathrm{Unknown}\right)=\frac{1-R\left(T\right)}{2}\end{array}\right.$

R in the formula (T), R (UT) and R (Unknown) constitute complete credit rating, R (T) expression mobile subscriber's confidence level, and R (UT) expression can not reliability, and R (Unknown) expression is uncertain;

For this calling subscriber calls out this mobile subscriber's average call duration, unit is second; T _IclBe all average call durations of going into to call out that this mobile subscriber accepts, unit is second; f _OcFor this mobile subscriber calls out this calling subscriber's calling frequency, unit is time/day; F _OcBe total average call frequency that this mobile subscriber calls out any user, unit is time/day; α and β are weight parameter.

22, method as claimed in claim 21 is characterized in that, this method also comprises the steps: portable terminal according to evaluation of user and call model parameter, produces credit rating according to following formula:

$\left\{\begin{array}{c}R\left(T\right)=\mathrm{\&alpha;}\frac{{\stackrel{\&OverBar;}{t}}_{\mathrm{icl}}}{{\stackrel{\&OverBar;}{T}}_{\mathrm{icl}}}+\mathrm{\&beta;}\frac{f_{\mathrm{oc}}}{{\stackrel{\&OverBar;}{F}}_{\mathrm{oc}}}\\ R\left(\mathrm{UT}\right)=[1-R\left(T\right)]\mathrm{\&gamma;}\\ R\left(\mathrm{Unknown}\right)=[1-R\left(T\right)](1-\mathrm{\&gamma;})\end{array}\right.$

R in the formula (T), R (UT) and R (Unknown) constitute complete credit rating, R (T) expression mobile subscriber's confidence level, and R (UT) expression can not reliability, and R (Unknown) expression is uncertain;

For this calling subscriber calls out this mobile subscriber's average call duration, unit is second; T _IclBe all average call durations of going into to call out that this mobile subscriber accepts, unit is second; f _OcFor this mobile subscriber calls out this calling subscriber's calling frequency, unit is time/day; F _OcBe total average call frequency that this mobile subscriber calls out any user, unit is time/day; α and β are weight parameter, and γ is this mobile subscriber's direct evaluation, and the value of γ is between 0～1.

23, as claim 21 or 22 described methods, it is characterized in that,, comprise and on average go into call duration and go out calling frequency so that mobile phone users is finished the credit rating to the calling subscriber according to the total call model parameter between this mobile subscriber and any user.

24, as claim 21 or 22 described methods, it is characterized in that weight parameter α and β can freely be defined by the mobile subscriber, make R (T) value between 0～1.

25, method as claimed in claim 22 is characterized in that, mobile phone users according to R (T), R (UT) and three opinion ratings of R (Unknown) can self-defined γ value.

26, as claim 21 or 22 described methods, it is characterized in that the credit rating that portable terminal provides adopts following form:

{R(T)，R(UT)，R(Unknown)}

In the formula, R (T) represents confidence level, the big more expression of R (T) calling subscriber is impossible more to be the rubbish voice user, R (UT) representative can not reliability, the big more expression of R (UT) calling subscriber may be the rubbish voice user more, R (Unknown) represents uncertainty, and the big more expression of R (Unknown) mobile subscriber can not determine more to calling subscriber's type.The value of R (T), R (UT) and R (Unknown) is all between 0～1, and three's sum is 1.

27, a kind of prestige appraisal procedure based on mobile radio communication is characterized in that this method comprises the steps:

The credit rating of mobile terminal receive begins the calling subscriber is carried out the prestige assessment;

If preserved this called subscriber's prestige set in the credit database of server, then the credit rating that this mobile phone users provided according to following formula of server is decayed and is obtained new credit rating:

$\left\{\begin{array}{c}R\&prime;\left(T\right)=R\left(T\right)\&times;\tilde{R}\left(T\right)\\ R\&prime;\left(\mathrm{UT}\right)=R\left(\mathrm{UT}\right)\&times;\tilde{R}\left(T\right)\\ R\&prime;\left(\mathrm{Unknown}\right)=1-R\&prime;\left(T\right)-R\&prime;\left(\mathrm{UT}\right)\end{array}\right.$

{ R ' (T), R ' (UT), R ' is (Unknown) } new credit rating set of expression in the formula, $\{\tilde{R}\left(T\right),\tilde{R}\left(\mathrm{UT}\right),\tilde{R}\left(\mathrm{Unknown}\right)\}$

Represent credit rating that user's historical prestige set is provided.

28, method as claimed in claim 27 is characterized in that, further utilizes the D-S evidence merging formula of evidence theory that the historical prestige of described new credit rating with this calling subscriber is merged.

29, method as claimed in claim 28 is characterized in that, further comprises described new credit rating set is reappraised to this calling subscriber's credit value according to following formula:

$\mathrm{Rep}=\frac{R\left(T\right)}{R\left(T\right)+R\left(\mathrm{Unknown}\right)}-\frac{R\left(\mathrm{UT}\right)}{R\left(\mathrm{UT}\right)+R\left(\mathrm{Unknown}\right)}$

Rep is this calling subscriber's a credit value in the formula, and { R (T), R (UT), R (Unknown) } is this calling subscriber's historical prestige set.

30, method as claimed in claim 29 is characterized in that, described calling subscriber's credit value Rep is between-1～+ 1.

31, method as claimed in claim 30 is characterized in that, comprises that further credit rating provides the user not have historical prestige set, and server is decayed to the credit rating of receiving with default prestige set, and default prestige set is set by server.

32, method as claimed in claim 30 is characterized in that, this targeted customer's historical prestige set is the result after being merged by all historical credit ratings, when receiving new credit rating, will reappraise and upgrade historical prestige set.

33, a kind of prestige based on mobile radio communication is shared method, it is characterized in that, from server download be lower than the prestige threshold values the rubbish voice user identifier to portable terminal, reach prestige and share.

34, method as claimed in claim 33 is characterized in that, described portable terminal requires to set different prestige threshold values according to the difference to rubbish voice filtering, and then downloads to the rubbish voice user identifier of varying number from server.

35, a kind of rubbish voice filtering system based on mobile radio communication, this system comprises: portable terminal and server is characterized in that described portable terminal comprises:

One man-machine interaction unit is used to set the rubbish voice credit value of portable terminal;

One rubbish voice user totem information storehouse is used to preserve the rubbish voice user identifier;

One white list is used to preserve the credible user identifier that the mobile subscriber once called out;

One mobile status monitoring unit is used for monitoring and goes into to call out and go out calling parameter;

One rubbish voice filtering unit filters the caller ID in the call request according to the rubbish voice user identifier in the described rubbish voice user totem information storehouse;

One rubbish voice user ID download unit is used for downloading up-to-date rubbish voice subscriber identifier information from server end;

One calling parameter statistic unit is monitored average duration and the calling frequency that calling parameter is added up the calling of all incoming calls of portable terminal and exhalation according to described mobile status monitoring unit;

One rubbish voice credit rating unit, the described parameter of adding up according to described calling parameter statistic unit is used for the calling subscriber who goes into to call out is produced credit rating;

One identity unit is used to finish the authentication of described mobile phone users at server end; And

One protocol communication unit is used to provide the function of the protocol communication between portable terminal and the server.

Described server end comprises:

One credit database is used for memory mobile terminal subscriber identity information and reputation data;

One prestige assessment unit carries out the prestige assessment according to new credit rating and the set of historical prestige that described called mobile terminal sends to the calling subscriber;

One reputation query unit is used for described inquiry credit database, finishes the reputation query requirement that portable terminal proposes; And

One communication unit is used for the protocol communication between server and the portable terminal.

36, system as claimed in claim 35 is characterized in that, described prestige assessment unit further comprises:

One credit rating attenuation module is used for the described credit rating that server receives is rationally decayed;

One credit rating merges module, is used for the described credit rating that server is received and the joint account of historical prestige set, obtains up-to-date prestige set;

One prestige quantization modules is calculated the credit value that is used to discern the rubbish voice user according to described prestige set.

37, system as claimed in claim 35 is characterized in that, described reputation query unit further comprises:

One real-time enquiry module is used to respond the real-time reputation query request that portable terminal proposes, and the prestige of the unique user of appointment is only inquired about in this request;

One non real-time download module is used to finish the rubbish voice user identifier download request that portable terminal proposes.

38, system as claimed in claim 35 is characterized in that, described rubbish voice user totem information storehouse further comprises:

One information stores management module is used for preserving and leading subscriber rubbish voice user identifier data, also is used for the index to the rubbish voice user identifier;

One message reference module is used for providing to rubbish voice filtering unit and rubbish voice credit rating unit the interface of visit information memory management unit, and this interface comprises inquiry, inserts and deletion.

39, system as claimed in claim 35 is characterized in that, described protocol communication unit further comprises:

One communication protocol module is used for the data that receive are decoded, and the data that send is encoded;

One message communicating module is used for transmitting and receive data between portable terminal and the server.

Images