CN107436877A - Much-talked-about topic method for pushing and device - Google Patents

Much-talked-about topic method for pushing and device Download PDF

Info

Publication number
CN107436877A
CN107436877A CN201610355128.9A CN201610355128A CN107436877A CN 107436877 A CN107436877 A CN 107436877A CN 201610355128 A CN201610355128 A CN 201610355128A CN 107436877 A CN107436877 A CN 107436877A
Authority
CN
China
Prior art keywords
mtd
topic
user
msub
mrow
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610355128.9A
Other languages
Chinese (zh)
Other versions
CN107436877B (en
Inventor
刘姗
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Original Assignee
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Jingdong Century Trading Co Ltd, Beijing Jingdong Shangke Information Technology Co Ltd filed Critical Beijing Jingdong Century Trading Co Ltd
Priority to CN201610355128.9A priority Critical patent/CN107436877B/en
Publication of CN107436877A publication Critical patent/CN107436877A/en
Application granted granted Critical
Publication of CN107436877B publication Critical patent/CN107436877B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/951—Indexing; Web crawling techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Abstract

This application discloses a kind of much-talked-about topic method for pushing and device.One embodiment of methods described includes：The comment information to topic that user delivers is obtained, and extracts the keyword of the topic；The comment information and the similarity of the keyword of extraction are calculated, similarity is more than the comment information of the first preset value as topic relevant information；Analysis made comments in the topic relevant information information each user between follow relation；Relation is followed based on described, calculates the value of the parameter preset of the topic, generates topic analysis report；Using topic corresponding to the topic analysis report for meeting the first preparatory condition as much-talked-about topic, the much-talked-about topic is pushed.The embodiment, which realizes, deeply comprehensively understands much-talked-about topic.

Description

Much-talked-about topic method for pushing and device
Technical field
The application is related to field of computer technology, more particularly to Internet technical field, and in particular to a kind of much-talked-about topic Method for pushing and device.
Background technology
With becoming increasingly popular for mobile Internet, Web Community turns into people and obtains information, the important channel of exchange of information. A large amount of netizens deliver the opinion of oneself in Web Community and disclose various news, have thousands of individual topics to produce daily.Such as What is chosen much-talked-about topic and pushed from so various topic, therefrom obtains the attitude to much-talked-about topic of netizen, will be right Solution social development situation, grasp the effect that public opinion trend plays directiveness.
At present, existing much-talked-about topic method for pushing mainly passes through some characteristic (e.g., nets in collection network community People's amount of posting, netizen geographic location etc. of posting), user or application program are pushed to after carrying out simple statistical analysis, no The comprehensive and deep understanding much-talked-about topic of user can be made.
The content of the invention
The purpose of the application is to propose a kind of much-talked-about topic method for pushing and device, to solve background section above The technical problem mentioned.
In a first aspect, this application provides a kind of much-talked-about topic method for pushing, methods described includes：Obtain what user delivered To the comment information of topic, and extract the keyword of the topic；It is similar to the keyword of extraction to calculate the comment information Degree, similarity is more than the comment information of the first preset value as topic relevant information；Analysis is in the topic relevant information Relation is followed between each user for information of making comments, and described to follow relation to include at least one of following：Reply relation, forwarding Relation, adduction relationship；Relation is followed based on described, calculates the value of the parameter preset of the topic, generates topic analysis report, institute Parameter preset is stated including at least one of following：Network clustering coefficient, network density, made comments in the topic relevant information The quantity of the user of information, wherein, the network clustering coefficient is by the every of information that made comments in the topic relevant information The cluster coefficients weighted average of individual user and obtain, the network density represents information of being made comments in the topic relevant information Each user the tightness degree for following relation, the network diameter is makes comments information in the topic relevant information The maximum of distance between two users；Topic corresponding to the topic analysis report for meeting the first preparatory condition is talked about as focus Topic, pushes the much-talked-about topic.
In certain embodiments, the similarity of the keyword for calculating the comment information and extraction, including：To described Comment information carries out word segmentation processing, obtains multiple words；Calculate the similarity of the multiple word and the keyword of extraction.
In certain embodiments, between each user for analyzing information of being made comments in the topic relevant information Relation is followed, including：When the comment information that user delivers in the topic relevant information is the reply topic relevant information In other users deliver comment information when, determine between two users to be reply relation；And/or when user is in the topic phase When closing the comment information that the comment information delivered in information is delivered for other users in the forwarding topic relevant information, two are determined It is forwarding relation between individual user；And/or when the comment information that user delivers in the topic relevant information is described in reference During the comment information that other users deliver in topic relevant information, determine between two users to be adduction relationship.
In certain embodiments, it is described to follow relation based on described, calculate the value of the parameter preset of the topic, generation words Analysis report is inscribed, including：Relation is followed based on described, establishing user using below equation follows relational matrix S：
Wherein, S is that the user follows relational matrix, and k is positive natural number, and N is to be delivered in the topic relevant information The quantity of the user of comment information, when S is that user replys relational matrix, kijReply what j-th of user delivered for i-th of user The number of comment information, when S is that user forwards relational matrix, kijThe comment for forwarding j-th of user to deliver for i-th of user is believed The number of breath, when S is user's adduction relationship matrix, kijTime for the comment information that j-th of user delivers is quoted for i-th of user Number, i ∈ [1, N], j ∈ [1, N], i ≠ j；As the kijDuring ＞ 0, i-th of user of line and j-th of user, build in the words The topological diagram for following relation of each user for information of being made comments in topic relevant information.
In certain embodiments, it is described to follow relation based on described, calculate the value of the parameter preset of the topic, generation words Analysis report is inscribed, including：Calculate each user's for information of being made comments in the topic relevant information according to below equation Cluster coefficients：
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in the topological diagram with i-th The neighbor user quantity of user's line, EiFor the wiring quantity between the neighbor user.
In certain embodiments, it is described to follow relation based on described, calculate the value of the parameter preset of the topic, generation words Analysis report is inscribed, including：The network density is calculated according to below equation：
Wherein, B is the network density, and L is the wiring quantity in the topological diagram.
In certain embodiments, it is described to follow relation based on described, calculate the value of the parameter preset of the topic, generation words Analysis report is inscribed, including：The network diameter is calculated according to below equation：
D=maxdij；
Wherein, D represents the network diameter, and d represents the distance in the topological diagram between each user, dijRepresent described Distance in topological diagram between i-th of user and j-th of user, max are represented to the dijTake maximum.
In certain embodiments, it is described to follow relation based on described, calculate the value of the parameter preset of the topic, generation words Analysis report is inscribed, including：The value of each parameter preset of the topic is weighted superposition, obtains comprehensive analysis parameter；Generation Include the topic analysis report of the comprehensive analysis parameter.
Second aspect, the application provide a kind of much-talked-about topic pusher, and described device includes：Comment information obtains single Member, the comment information to topic delivered for obtaining user, and extract the keyword of the topic；Relevant information determines single Member, for calculating the comment information and the similarity of the keyword of extraction, the comment that similarity is more than to the first preset value is believed Breath is used as topic relevant information；Relationship analysis unit is followed, for analyzing information of being made comments in the topic relevant information Each user between follow relation, it is described that to follow relation to include at least one of following：Reply relation, forwarding relation, quote and close System；Analysis report generation unit, for following relation based on described, the value of the parameter preset of the topic is calculated, generates topic Analysis report, the parameter preset include at least one of following：Cluster coefficients, network density, network diameter, in the topic phase The quantity of the user for information of being made comments in information is closed, wherein, the cluster coefficients in the topic relevant information by delivering The cluster coefficients weighted average of each user of comment information and obtain, the network density is represented in the topic relevant information The tightness degree for following relation of each user for information of making comments, the network diameter are to be sent out in the topic relevant information The maximum of distance between two users of table comment information；Much-talked-about topic push unit, for the first preparatory condition will to be met Topic analysis report corresponding to topic as much-talked-about topic, push the much-talked-about topic.
In certain embodiments, the relevant information determining unit, including：Word-dividing mode, for the comment information Word segmentation processing is carried out, obtains multiple words；Computing module, it is similar to the keyword of extraction for calculating the multiple word Degree.
In certain embodiments, it is described to follow relationship analysis unit, including：Determining module is replied, for when user is in institute State the comment information that the comment information delivered in topic relevant information is delivered for other users in the reply topic relevant information When, determine between two users to be reply relation；And/or forwarding determining module, for when user is in the topic relevant information In the comment information delivered to forward other users deliver in the topic relevant information comment information when, determine two users Between be forwarding relation；And/or determining module is quoted, for the comment information delivered as user in the topic relevant information During to quote other users deliver in the topic relevant information comment information, determine between two users to be adduction relationship.
In certain embodiments, the analysis report generation unit, in addition to：Follow relational matrix to establish module, be used for Relation is followed based on described, establishing user using below equation follows relational matrix S：
Wherein, S is that the user follows relational matrix, and k is positive natural number, and N is to be delivered in the topic relevant information The quantity of the user of comment information, when S is that user replys relational matrix, kijReply what j-th of user delivered for i-th of user The number of comment information, when S is that user forwards relational matrix, kijThe comment for forwarding j-th of user to deliver for i-th of user is believed The number of breath, when S is user's adduction relationship matrix, kijTime for the comment information that j-th of user delivers is quoted for i-th of user Number, i ∈ [1, N], j ∈ [1, N], i ≠ j；Topological diagram builds module, for as the kijDuring ＞ 0, i-th of user of line and jth Individual user, build the topological diagram for following relation of each user for information of being made comments in the topic relevant information.
In certain embodiments, the analysis report generation unit, in addition to：Cluster coefficients computing module, for basis Below equation calculates the cluster coefficients of each user for information of being made comments in the topic relevant information：
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in the topological diagram with i-th The neighbor user quantity of user's line, EiFor the wiring quantity between the neighbor user.
In certain embodiments, the analysis report generation unit, in addition to：Network density computing module, for basis Below equation calculates the network density：
Wherein, B is the network density, and L is the wiring quantity in the topological diagram.
In certain embodiments, the analysis report generation unit, in addition to：Network diameter computing module, for basis Below equation calculates the network diameter：
D=maxdij；
Wherein, D represents the network diameter, and d represents the distance in the topological diagram between each user, dijRepresent described Distance in topological diagram between i-th of user and j-th of user, max are represented to the dijTake maximum.
In certain embodiments, the analysis report generation unit, in addition to：Comprehensive analysis parameter determination module, is used for The value of each parameter preset of the topic is weighted superposition, obtains comprehensive analysis parameter；Topic analysis report generation module, Include the topic analysis report of the comprehensive analysis parameter for generating.
The much-talked-about topic method for pushing and device that the application provides, by extracting the comment information of topic, and pass through calculating The mode of similarity chooses topic relevant information, analyze between the user for information of being made comments in above-mentioned topic relevant information with Relation is followed with relation, and according to above-mentioned, the value of the parameter preset of topic is calculated, obtains topic analysis report, is then chosen full Topic corresponding to the topic analysis report of the first preparatory condition of foot as much-talked-about topic and is pushed, so as to make user more It is deeply comprehensive to understand much-talked-about topic.
Brief description of the drawings
By reading the detailed description made to non-limiting example made with reference to the following drawings, the application's is other Feature, objects and advantages will become more apparent upon：
Fig. 1 is that the application can apply to exemplary system architecture figure therein；
Fig. 2 is the flow chart according to one embodiment of the much-talked-about topic method for pushing of the application；
Fig. 3 is the flow chart according to the parameter preset of the calculating topic of the much-talked-about topic method for pushing of the application；
Fig. 4 is to follow relationship topology figure according to the user of one embodiment of the much-talked-about topic method for pushing of the application；
Fig. 5 is the calculation process according to the network clustering coefficient of one embodiment of the much-talked-about topic method for pushing of the application Figure；
Fig. 6 is the structural representation according to one embodiment of the much-talked-about topic pusher of the application；
Fig. 7 is adapted for the structural representation of the computer system of the server for realizing the embodiment of the present application.
Embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to Be easy to describe, illustrate only in accompanying drawing to about the related part of invention.
It should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the application can phase Mutually combination.Describe the application in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 shows the embodiment of the much-talked-about topic method for pushing that can apply the application or much-talked-about topic pusher Exemplary system architecture 100.
As shown in figure 1, system architecture 100 can include terminal device 101,102,103, network 104 and server 105. Network 104 between terminal device 101,102,103 and server 105 provide communication link medium.Network 104 can be with Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be interacted with using terminal equipment 101,102,103 by network 104 with server 105, to receive or send out Send message etc..Various telecommunication customer end applications can be installed, such as web browser should on terminal device 101,102,103 With, searching class application, JICQ, mailbox client, social platform software etc..
Terminal device 101,102,103 can have display screen and supported web page browses or the various electricity of social communication Sub- equipment, including but not limited to smart mobile phone, tablet personal computer, E-book reader, pocket computer on knee and desk-top calculating Machine etc..
Server 105 can be to provide the server of various services, such as to being shown on terminal device 101,102,103 Webpage or social platform provide the background server supported.Background server can divide data such as the comment informations of acquisition The processing such as analysis, and result (such as much-talked-about topic) is fed back into terminal device.
It should be noted that the much-talked-about topic method for pushing that the embodiment of the present application is provided typically is performed by server 105, Correspondingly, much-talked-about topic pusher is generally positioned in server 105.
It should be understood that the number of the terminal device, network and server in Fig. 1 is only schematical.According to realizing need Will, can have any number of terminal device, network and server.
With continued reference to Fig. 2, the flow chart of one embodiment of much-talked-about topic method for pushing according to the application is shown 200.The much-talked-about topic method for pushing of the present embodiment comprises the following steps：
Step 201, the comment information to topic that user delivers is obtained, and extracts the keyword of above-mentioned topic.
In the present embodiment, it can pass through terminal (such as terminal device shown in Fig. 1) point in user to obtain comment information Some for performing after hitting the specified location of some application of installation in the terminal or being installed on user opens a terminal should Used time performs.Also, it is understood that user is in the application program installed on clicking on or opening a terminal, to server (example Server as shown in Figure 1) send much-talked-about topic push request.
In the present embodiment, electronic equipment (such as the service shown in Fig. 1 of much-talked-about topic method for pushing operation thereon Device) end of web page browsing or social communication can be carried out using it from user by wired connection mode or radio connection End receives much-talked-about topic push request.It is pointed out that above-mentioned radio connection can include but is not limited to 3G/4G companies Connect, WiFi connections, bluetooth connection, WiMAX connections, Zigbee connections, UWB (ultra wideband) connections and other are existing In known or exploitation in the future radio connection.
In the present embodiment, the comment information to topic that user delivers can be the registered user or non-registered of any website The comment information that user delivers below topic.It is understood that in the present embodiment, nonregistered user can be earnest net Friend or the user logged in by third party software.The form of expression of comment information can include：Word, character expression, animation table Feelings, picture with word etc. can express any form of expression to the viewpoint of topic.The form of expression of topic can be microblogging Or the form of expression of the topic in webpage, such as # Beijing haze #, vagrant dog guards old man, etc. form.In the present embodiment, carry Take the keyword of topic, using obtained each word as keyword, can also pass through spy by being segmented to topic Fixed analysis mode (such as statistical analysis mode, semantic analysis mode) determines keyword, can also be using topic in itself as key Word.Obtaining above-mentioned comment information can obtain by way of reptile crawls.
Generally, user using the web browser or network comment community application program installed in terminal come in webpage Topic is made comments information.In the present embodiment, above-mentioned webpage can include html forms, xhtml forms, asp forms, php Form, jsp forms, shtml forms, nsp forms, the webpage of xml forms or other following forms by exploitation webpage (only The contents such as picture, animation, word that it includes can be opened and browse with browser by wanting the web page files of this form).
Step 202, comment information and the similarity of the keyword of extraction are calculated, similarity is more than commenting for the first preset value By information as topic relevant information.
In the present embodiment, Similarity Measure algorithm can be used to the similarity between comment information and the keyword of extraction Calculated, it is for instance possible to use cosine similarity (cosine similarity) algorithm, simhash algorithms etc is known Text similarity computing method carry out Similarity Measure.
In some optional embodiments of the present embodiment, step 202 specifically includes the following sub-step not shown in Fig. 2 Suddenly：
Word segmentation processing is carried out to comment information, obtains multiple words；Calculate the multiple word and the keyword of extraction Similarity.
In the present embodiment, word segmentation processing is carried out to comment information, comment information can be cut by the way of full cutting Into word.
By taking simhash algorithms as an example, it passes through the Hamming distances (Hamming of the simhash values of more multiple documents Distance), their similarity can be obtained.Specific work process refers to as follows：Calculate the keyword of extraction Simhash values, store it in the internal memory of server.By every comment information to be compared simhash values corresponding with its It is carried in redis databases, by simhash values corresponding to comment information to be compared and the pass for the extraction being stored in internal memory The simhash values of keyword are compared, and obtain the Hamming distances of two simhash values.Hamming distances are more than the first preset value Comment information as topic relevant information.
In the present embodiment, the specific value of the first preset value is not defined, those skilled in the art can be according to reality Need the value of the preset value of sets itself first.
Step 203, analyze and follow relation between each user for information of being made comments in topic relevant information.
It is above-mentioned to follow relation be including at least one in following relation in the present embodiment：Reply relation, forwarding are closed System, adduction relationship.
In some optional implementations of the present embodiment, step 203 specifically includes the following sub-step not shown in Fig. 2 Suddenly：
Sub-step one, it is other in topic relevant information to reply when the comment information that user delivers in topic relevant information During the comment information that user delivers, determine between two users to be reply relation；And/or
Sub-step two, it is other in the comment information that user delivers in topic relevant information is forwarding topic relevant information During the comment information that user delivers, determine between two users to be forwarding relation；And/or
Sub-step three, it is other in topic relevant information to quote when the comment information that user delivers in topic relevant information During the comment information that user delivers, determine between two users to be adduction relationship.
For convenience of understanding, this sentences the party A-subscriber for information of being made comments in topic relevant information and party B-subscriber two is different Exemplified by user, carry out the above-mentioned three kinds of relations of specific explanations.Reply relation can be understood as：Party A-subscriber is delivering a comment information Afterwards, party B-subscriber is replied for the comment information that party A-subscriber delivers, then is reply relation between party A-subscriber and party B-subscriber.Forwarding relation It can be understood as：After party A-subscriber has delivered a comment information, party B-subscriber forwarded the comment information that party A-subscriber delivers, then party A-subscriber and It is forwarding relation between party B-subscriber.Adduction relationship can be understood as：After party A-subscriber has delivered a comment information, party B-subscriber refer to A The comment information that user delivers, then it is adduction relationship between party A-subscriber and party B-subscriber.
Step 204, based on relation is followed, the value of the parameter preset of above-mentioned topic is calculated, generates topic analysis report.
In the present embodiment, parameter preset includes at least one of following：Cluster coefficients, network density, network diameter, in topic The quantity of the user for information of being made comments in relevant information.Wherein, cluster coefficients are by letter of being made comments in topic relevant information The cluster coefficients weighted average of each user of breath and obtain, network density expression is made comments information in topic relevant information The tightness degree for following relation of each user.Network density is bigger, represent user between follow relation closer, and user it Between reply or the number of reference or forwarding it is more.Network diameter is two use of information of being made comments in topic relevant information The maximum of distance between family.
In the present embodiment, topic analysis report can include value, the value according to above-mentioned parameter preset of each parameter preset One obtained is used for the value for reflecting topic temperature or the analysis result being worth to for above-mentioned parameter preset etc..
In some optional implementations of the present embodiment, step 204 specifically includes the following sub-step not shown in Fig. 2 Suddenly：
The value of each parameter preset of above-mentioned topic is weighted superposition, obtains comprehensive analysis parameter；Generation includes synthesis The topic analysis report of analytical parameters.
In the present embodiment, different weights can be set for different parameter presets in advance, the value of each parameter preset is entered Row weighted superposition obtains above-mentioned comprehensive analysis parameter, and generation includes the topic analysis report of comprehensive analysis parameter.
Step 205, using topic corresponding to the topic analysis report for meeting the first preparatory condition as much-talked-about topic, push heat Point topic.
In the present embodiment, the first preparatory condition can be the threshold value array set for each parameter preset, i.e., each pre- When the value of setting parameter reaches corresponding threshold value in threshold value array, it is believed that this topic is much-talked-about topic.First preparatory condition can be with Set according to analysis result, for example, result be temperature it is high when, using corresponding topic as much-talked-about topic.It is understood that the One preparatory condition is the content in topic analysis report and set.It is determined that after much-talked-about topic, much-talked-about topic is pushed The user of webpage to the application program installed in terminal or is being browsed by the web browser in terminal.
The much-talked-about topic method for pushing that the present embodiment provides, by extracting the comment information of topic, and it is similar by calculating The mode of degree chooses topic relevant information, analyzes and follows pass between the user for information of being made comments in above-mentioned topic relevant information System, and follows relation according to above-mentioned, calculates the value of the parameter preset of topic, obtains topic analysis report, then chooses and meets the Topic corresponding to the topic analysis report of one preparatory condition as much-talked-about topic and is pushed, so as to make user deeper into It is comprehensive to understand much-talked-about topic.
With further reference to Fig. 3, the parameter preset of the calculating topic according to the much-talked-about topic method for pushing of the application is shown Flow chart 300.The much-talked-about topic method for pushing of the present embodiment is by using methods of social network (Social Network Analysis, SNA) study the relation in topic relevant information between each comment information.Social network analysis is one group of row of research The research method of the relation of dynamic person.One group of actor can be people, community, colony, tissue, country etc., their relation schema The phenomenon or data that reflect are the focuses of network analysis.In the present embodiment, using actor and its mutual relation as grinding Study carefully content, by the way that the relation between actor is described, analyze structure that these relations are contained and its to actor and The influence of whole colony.
During the analysis of the much-talked-about topic of the present embodiment, using in network each user as node, each user it Between the relation that follows be line.So network can be described with the line set between node combination and node.
The parameter preset of the calculating topic of the much-talked-about topic method for pushing of the present embodiment comprises the following steps：
Step 301, based on relation is followed, establish user and follow relational matrix.
In the present embodiment, with SNA methods, user is established using below equation and follows relational matrix S：
Wherein, S is that user follows relational matrix, and k is positive natural number, and N is information of being made comments in topic relevant information User quantity, when S be user reply relational matrix when, kijThe comment information delivered for i-th of user j-th of user of reply Number, when S be user forward relational matrix when, kijTime for the comment information delivered for i-th of user j-th of user of forwarding Number, when S is user's adduction relationship matrix, kijThe number for the comment information that j-th of user delivers, i ∈ are quoted for i-th of user [1, N], j ∈ [1, N], i ≠ j.
Step 302, the topology for following relation of each user for information of being made comments in the topic relevant information is built Figure.
In user obtained above follows relational matrix, work as kijDuring ＞ 0, i-th of user of line and j-th of user, it is Become apparent from intuitively showing and follow relation between user, a topological diagram can be built follow pass specifically to represent above-mentioned System.
It is specifically described exemplified by reply relation between user.Assuming that topic relevant information includes 6 users, point Wei not user a, b, c, d, e and f.Use as shown in Figure 4 is established according to the reply relation of above-mentioned 6 users in topic relevant information Reply relationship topology figure in family.Oriented arrow in Fig. 4 between user a and user b represents that user b has replied user a.User a with Oriented arrow between user c represents to be replied mutually therebetween.By Fig. 4 topological diagram it can also be seen that user a be The user of starting comment information is delivered in topic relevant information.
It is understood that the topological diagram between each user for information of being made comments in topic relevant information is built When, it can first determine to deliver the user of starting comment information, i.e. originating subscriber in topic relevant information.In practice, network In community, the comment information that user delivers exists in the form of model.Therefore, above-mentioned starting comment information is found to can be understood as Find first note.In Web Community, each model has fileID (file identifier) to be used to represent it in topic relevant information In position.First note can be determined according to fileID, then looks for use corresponding with other models that first note has reply relation Family, referred to as neighbor user.The neighbor user that reply relation be present with above-mentioned neighbor user is determined again, with line segment between neighbor user To connect, so as to build the topological diagram for following relation of each user for information of being made comments in topic relevant information.
Step 3031, cluster coefficients are calculated.
Cluster coefficients are to weigh the important parameter of network group.Assuming that a node i in network has liBar side by it It is connected with other nodes, this liIndividual node just turns into the neighbor node of node i.In relational network, certain two of a node It is likely to also be relatively neighbours between neighbor node, this attribute is referred to as the Clustering features of network.In theory, the l of node iiIt is individual May at most there is l between neighbor nodei(li- 1)/2 side, and this liThe side number E of physical presence between individual nodeiCan with total The side number l of energyi(li- 1)/the ratio between 2 just it is defined as the cluster coefficients C of node ii.Thus, in the present embodiment, according to below equation meter Calculate the cluster coefficients of each user for information of being made comments in topic relevant information：
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in the topological diagram with i-th The neighbor user quantity of user's line, EiFor the wiring quantity between above-mentioned neighbor user.
Network clustering coefficient by information of being made comments in topic relevant information each user cluster coefficients weighted average And obtain.Then, network clustering coefficient can be calculated by below equation：
Wherein, C is network clustering coefficient, CiFor the cluster coefficients of i-th of user, N is in the topic relevant information The quantity of the user for information of making comments, i ∈ [1, N].
Step 3032, calculating network density.
Network density represent made comments in topic relevant information information each user the tightness degree for following relation. In the present embodiment, according to below equation calculating network density：
Wherein, B is network density, and L is the wiring quantity in topological diagram.
Step 3033, calculating network diameter.
Network diameter can reflect the size of topological diagram, and in the present embodiment, introducing network diameter can be used for reflecting topic The range size of influence, the density degree of topological diagram interior joint (user) contact can also be reflected.Any two in SNA methods The maximum of the distance between node is referred to as network diameter.Thus, in the present embodiment, it is straight that the network is calculated according to below equation Footpath：
D=maxdij；
Wherein, D represents network diameter, and d represents the distance in topological diagram between each user, dijRepresent i-th in topological diagram Distance between individual user and j-th of user, max are represented to dijTake maximum.
Step 304, comprehensive analysis parameter is calculated, generates topic analysis report.
The much-talked-about topic method for pushing that the present embodiment provides, by following the analysis of relation between the user to topic, lead to The value for calculating parameter preset is crossed, obtains topic analysis report, then the much-talked-about topic of selection is pushed to user, can effectively be helped Helping user, deep understanding much-talked-about topic and netizen to the attitude of much-talked-about topic, to understanding spin there is directiveness to make comprehensively With.
The calculating process of network clustering coefficient can show the much-talked-about topic push side according to the application with specific reference to Fig. 5 The calculation flow chart 500 of the network clustering coefficient of one embodiment of method.The calculating of the network clustering coefficient of the present embodiment includes Following steps：
Step 501, the descending sequence of quantity for the comment information delivered by user in topic relevant information.
Because the number of users for information of being made comments in Web Community is huge, comment information quantity that some users deliver Less, its corresponding cluster coefficients is just very small, plays a part of in calculating network cluster coefficients limited.Calculated to improve Efficiency, the cluster coefficients for often choosing a part of user are counted.In the present embodiment, information content of making comments is chosen Larger user is as calculating object.
First, user descending is ranked up according to the quantity for information of being made comments in topic relevant information.
Step 502, to each user in preceding M user, determine that the user delivers all in topic relevant information Comment information.
Using the M maximum user for information of being made comments in topic relevant information as calculating object.For above-mentioned M Each user of user, determine all comment informations that the user delivers in topic relevant information.It is understood that M Any positive natural number of the desirable number of users for being less than information of being made comments in topic relevant information.
Step 503, the neighbor user collection unification for all users that the user replys is determined.
All comment informations delivered according to the user, find all users of user reply, and these users are referred to as The neighbor user of the user, its collection formed are collectively referred to as the unification of neighbor user collection.
Step 504, it is determined that replying the neighbor user set two of all users of the user.
It can be found according to the mark (such as account title) of the user in topic relevant information and reply the user's All users, these users are also the neighbor user of the user.The collection that these neighbor users are formed is collectively referred to as neighbor user set Two.
Step 505, merge collection unification and set two for set three, the duplicate customer in set three is removed, it is determined that in topology The session number E of the user in figureA。
Neighbor user collection is unified and may include some identical neighbor users in neighbor user set two, and these neighbours use Family has Double-directional back returning to customs system with the user.Merging collection unification and set two, after obtaining set three, these in set three have The neighbor user of Double-directional back returning to customs system is calculated twice, therefore to remove three kinds of these duplicate customers of set.Structure should simultaneously The topological diagram of user, determine the wiring quantity E of the user in topological diagramA。
Step 506, for each neighbor user in set three, the set for all users that the neighbor user is replied is determined Four.
All users that neighbor user is replied are the neighbor user of neighbor user, are sake of clarity, can be referred to as two herein Level neighbor user.But the two level neighbor user of neighbor user reply is looked only for herein.
Step 507, when other neighbor users in set three belong to set four, EAAdd 1.
When other neighbor users in set three belong to set four, illustrate all to deposit between user and two neighbor user Reply relation, now session number of the user in topological diagram to add 1.
Step 508, the duplicate customer in user's set that all neighbor users in set three are replied is removed.
As an example it is assumed that set three includes 4 neighbor users, its user replied set is respectively that two level neighbours use Family set I and II neighbor user set two, two level neighbor user set three and two level neighbor user set four, this four two levels The user of repetition is there may be in neighbor user set, therefore to remove these users repeated.
Step 509, before calculating in M user each user cluster coefficients.
The quantity of each wiring quantity and user of the user in topological diagram in preceding M user is determined, can calculate correspondingly Cluster coefficients.
Step 510, calculating network cluster coefficients.
Network clustering coefficient is obtained by the cluster coefficients weighted average calculation of preceding M user.
The much-talked-about topic method for pushing that the present embodiment provides, passes through the user for information of being made comments in topic relevant information It is middle to choose representational user, and careful calculating is carried out to the cluster coefficients of these representative users, obtain network clustering system Number, the relation that follows that can be sufficiently reflected between the user for information of being made comments in topic relevant information, deep enough can grind The temperature of topic is studied carefully, so as to push accurate much-talked-about topic.
With further reference to Fig. 6, show that the structure of one embodiment of the much-talked-about topic pusher according to the application is shown It is intended to 600.As shown in fig. 6, the much-talked-about topic pusher of the present embodiment includes：Comment information acquiring unit 601, relevant information Determining unit 602, follow relationship analysis unit 603, analysis report generation unit 604 and much-talked-about topic push unit 605.
Wherein, comment information acquiring unit 601, the comment information to topic delivered for obtaining user, and extract words The keyword of topic.In the present embodiment, the comment information to topic that user delivers in network can be the registration use of any website The comment information that family or nonregistered user are delivered below topic.In the present embodiment, extracting the keyword of topic can pass through Participle is carried out to topic and determines keyword, keyword can also be determined by specific analysis mode, can also be by topic in itself As keyword.
Relevant information determining unit 602, the keyword extracted for calculating comment information with comment information acquiring unit 601 Similarity, using similarity be more than the first preset value comment information as topic relevant information.In the present embodiment, it can use Similarity Measure algorithm is calculated the similarity between comment information and the keyword of extraction.
Relationship analysis unit 603 is followed, for analyzing in the topic relevant information that relevant information determining unit 602 determines Relation is followed between each user for information of making comments, follows relation to include at least one of following：Reply relation, forwarding are closed System, adduction relationship.
Analysis report generation unit 604, for based on following relationship analysis unit 603 to analyze the obtained relation that follows, counting The value of the parameter preset of topic is calculated, generates topic analysis report, the parameter preset includes at least one of following：Cluster coefficients, Network density, network diameter, information of making comments in topic relevant information user quantity, wherein, cluster coefficients by The cluster coefficients weighted average of each user for information of being made comments in topic relevant information and obtain, network density represent in topic The tightness degree for following relation between each user for information of being made comments in relevant information.Network diameter is letter related in topic The maximum of distance between two users of information of being made comments in breath
Much-talked-about topic push unit 605, for meeting in the topic analysis report that obtains analysis report generation unit 604 Topic corresponding to the topic analysis report of first preparatory condition pushes much-talked-about topic as much-talked-about topic.It is determined that much-talked-about topic Afterwards, much-talked-about topic is pushed to the application program installed in terminal or webpage is browsed by the web browser in terminal User.
In some optional implementations of the present embodiment, relevant information determining unit 602, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Word-dividing mode, for carrying out word segmentation processing to comment information, obtain multiple words；Computing module is more for calculating The similarity of individual word and the keyword of extraction.
Obtained multiple words are sent to computing module to calculate similarity by word-dividing mode.
In some optional implementations of the present embodiment, follow relationship analysis unit 603 to specifically include in Fig. 6 and do not show Go out with lower module：
Determining module is replied, the comment information for being delivered as user in topic relevant information is believed to reply topic correlation During the comment information that other users deliver in breath, determine between two users to be reply relation；And/or
Determining module is forwarded, the comment information for being delivered as user in topic relevant information is the related letter of forwarding topic During the comment information that other users deliver in breath, determine between two users to be forwarding relation；And/or
Determining module is quoted, the comment information for being delivered as user in topic relevant information is believed to quote topic correlation During the comment information that other users deliver in breath, determine between two users to be adduction relationship.
In some optional implementations of the present embodiment, analysis report generation unit 604, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Follow relational matrix to establish module, for following relation based on above-mentioned, user is established using below equation and follows pass It is matrix S：
Wherein, S is that the user follows relational matrix, and k is positive natural number, and N is to be delivered in the topic relevant information The quantity of the user of comment information, when S is that user replys relational matrix, kijReply what j-th of user delivered for i-th of user The number of comment information, when S is that user forwards relational matrix, kijThe comment for forwarding j-th of user to deliver for i-th of user is believed The number of breath, when S is user's adduction relationship matrix, kijTime for the comment information that j-th of user delivers is quoted for i-th of user Number, i ∈ [1, N], j ∈ [1, N], i ≠ j.
In order to follow relation between cheer and bright reflection user, topological diagram can be built for further calculating.
Topological diagram builds module, for as the kijDuring ＞ 0, i-th of user of line and j-th of user, build in topic The topological diagram for following relation of each user for information of being made comments in relevant information.
In some optional implementations of the present embodiment, analysis report generation unit 604, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Cluster coefficients computing module, for calculating the every of information that made comments in topic relevant information according to below equation The cluster coefficients of individual user：
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in topological diagram with i-th of user The neighbor user quantity of line, EiFor the wiring quantity between above-mentioned neighbor user.
In some optional implementations of the present embodiment, analysis report generation unit 604, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Network density computing module, for according to below equation calculating network density：
Wherein, B is network density, and L is the wiring quantity in topological diagram.
In some optional implementations of the present embodiment, analysis report generation unit 604, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Network diameter computing module, for calculating the network diameter according to below equation：
D=maxdij；
Wherein, D represents network diameter, and d represents the distance in topological diagram between each user, dijRepresent i-th in topological diagram Distance between individual user and j-th of user, max are represented to dijTake maximum.
In some optional implementations of the present embodiment, analysis report generation unit 604, which is specifically included in Fig. 6, not to be shown Go out with lower module：
Comprehensive analysis parameter determination module, for the value of each parameter preset of topic to be weighted into superposition, integrated Analytical parameters；Topic analysis report generation module, the topic analysis report of above-mentioned comprehensive analysis parameter is included for generating.
The much-talked-about topic pusher that the present embodiment provides, by extracting the comment information of topic, and it is similar by calculating The mode of degree chooses topic relevant information, analyzes and follows pass between the user for information of being made comments in above-mentioned topic relevant information System, and follows relation according to above-mentioned, calculates the value of the parameter preset of topic, obtains topic analysis report, then chooses and meets the Topic corresponding to the topic analysis report of one preparatory condition as much-talked-about topic and is pushed, so as to make user deeper into It is comprehensive to understand much-talked-about topic.
Below with reference to Fig. 7, it illustrates suitable for for realizing the computer system 700 of the server of the embodiment of the present application Structural representation.
As shown in fig. 7, computer system 700 includes CPU (CPU) 701, it can be read-only according to being stored in Program in memory (ROM) 702 or be loaded into program in random access storage device (RAM) 703 from storage part 708 and Perform various appropriate actions and processing.In RAM 703, also it is stored with system 700 and operates required various programs and data. CPU 701, ROM 702 and RAM 703 are connected with each other by bus 704.Input/output (I/O) interface 705 is also connected to always Line 704.
I/O interfaces 705 are connected to lower component：Importation 706 including keyboard, mouse etc.；Penetrated including such as negative electrode The output par, c 707 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage part 708 including hard disk etc.； And the communications portion 709 of the NIC including LAN card, modem etc..Communications portion 709 via such as because The network of spy's net performs communication process.Driver 710 is also according to needing to be connected to I/O interfaces 705.Detachable media 711, such as Disk, CD, magneto-optic disk, semiconductor memory etc., it is arranged on as needed on driver 710, in order to read from it Computer program be mounted into as needed storage part 708.
Especially, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description Software program.For example, embodiment of the disclosure includes a kind of computer program product, it includes being tangibly embodied in machine readable Computer program on medium, the computer program include the program code for being used for the method shown in execution flow chart.At this In the embodiment of sample, the computer program can be downloaded and installed by communications portion 709 from network, and/or from removable Medium 711 is unloaded to be mounted.When the computer program is performed by CPU (CPU) 701, perform in the present processes The above-mentioned function of limiting.
Flow chart and block diagram in accompanying drawing, it is illustrated that according to the system of the various embodiments of the application, method and computer journey Architectural framework in the cards, function and the operation of sequence product.At this point, each square frame in flow chart or block diagram can generation The part of one module of table, program segment or code, a part for the module, program segment or code include one or more For realizing the executable instruction of defined logic function.It should also be noted that some as replace realization in, institute in square frame The function of mark can also be with different from the order marked in accompanying drawing generation.For example, two square frames succeedingly represented are actual On can perform substantially in parallel, they can also be performed in the opposite order sometimes, and this is depending on involved function.Also It is noted that the combination of each square frame and block diagram in block diagram and/or flow chart and/or the square frame in flow chart, Ke Yiyong Function as defined in execution or the special hardware based system of operation are realized, or can be referred to specialized hardware and computer The combination of order is realized.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard The mode of part is realized.Described unit can also be set within a processor, for example, can be described as：A kind of much-talked-about topic Pusher includes：Comment information acquiring unit, relevant information determining unit, follow relationship analysis unit, analysis report generation Unit and much-talked-about topic push unit.Wherein, the title of these units is not formed to the unit in itself under certain conditions Limit, for example, comment information acquiring unit is also described as " obtaining the comment information to topic that user delivers, and carrying Take the unit of the keyword of the topic ".
As on the other hand, present invention also provides a kind of nonvolatile computer storage media, the non-volatile calculating Machine storage medium can be the nonvolatile computer storage media included in device described in above-described embodiment；Can also be Individualism, without the nonvolatile computer storage media in supplying terminal.Above-mentioned nonvolatile computer storage media is deposited One or more program is contained, when one or more of programs are performed by an equipment so that the equipment：Obtain The comment information to topic that user delivers, and extract the keyword of the topic；Calculate the comment information and the pass of extraction The similarity of keyword, similarity is more than the comment information of the first preset value as topic relevant information；Analysis is in the topic Relation is followed between each user for information of being made comments in relevant information, and described to follow relation to include at least one of following：Return Returning to customs system, forwarding relation, adduction relationship；Relation is followed based on described, calculates the value of the parameter preset of the topic, generates topic Analysis report, the parameter preset include：Cluster coefficients, network density, network diameter, delivered in the topic relevant information The quantity of the user of comment information, wherein, the cluster coefficients are by the every of information that made comments in the topic relevant information The cluster coefficients weighted average of individual user and obtain, the network density represents information of being made comments in the topic relevant information Each user the tightness degree for following relation, the network diameter is makes comments information in the topic relevant information The maximum of distance between two users；Topic corresponding to the topic analysis report for meeting the first preparatory condition is talked about as focus Topic, pushes the much-talked-about topic.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.People in the art Member should be appreciated that invention scope involved in the application, however it is not limited to the technology that the particular combination of above-mentioned technical characteristic forms Scheme, while should also cover in the case where not departing from the inventive concept, carried out by above-mentioned technical characteristic or its equivalent feature The other technical schemes for being combined and being formed.Such as features described above has similar work(with (but not limited to) disclosed herein The technical scheme that the technical characteristic of energy is replaced mutually and formed.

Claims (16)

1. a kind of much-talked-about topic method for pushing, it is characterised in that methods described includes：
The comment information to topic that user delivers is obtained, and extracts the keyword of the topic；
The comment information and the similarity of the keyword of extraction are calculated, the comment information that similarity is more than to the first preset value is made For topic relevant information；
Analysis made comments in the topic relevant information information each user between follow relation, it is described to follow relation bag Include at least one of following：Reply relation, forwarding relation, adduction relationship；
Relation is followed based on described, calculates the value of the parameter preset of the topic, generates topic analysis report, the parameter preset Including at least one of following：Network clustering coefficient, network density, network diameter, made comments in the topic relevant information The quantity of the user of information, wherein, the network clustering coefficient is by the every of information that made comments in the topic relevant information The cluster coefficients weighted average of individual user and obtain, the network density represents information of being made comments in the topic relevant information Each user the tightness degree for following relation, the network diameter is makes comments information in the topic relevant information The maximum of distance between two users；
Using topic corresponding to the topic analysis report for meeting the first preparatory condition as much-talked-about topic, the much-talked-about topic is pushed.
2. according to the method for claim 1, it is characterised in that described to calculate the comment information and the keyword of extraction Similarity, including：
Word segmentation processing is carried out to the comment information, obtains multiple words；
Calculate the similarity of the multiple word and the keyword of extraction.
3. according to the method for claim 1, it is characterised in that the analysis is made comments in the topic relevant information Relation is followed between each user of information, including：
Other users in the comment information that user delivers in the topic relevant information is the reply topic relevant information During the comment information delivered, determine between two users to be reply relation；And/or
Other users in the comment information that user delivers in the topic relevant information is the forwarding topic relevant information During the comment information delivered, determine between two users to be forwarding relation；And/or
Other users in the comment information that user delivers in the topic relevant information is the reference topic relevant information During the comment information delivered, determine between two users to be adduction relationship.
4. according to the method for claim 3, it is characterised in that it is described to follow relation based on described, calculate the topic The value of parameter preset, topic analysis report is generated, including：
Relation is followed based on described, establishing user using below equation follows relational matrix S：
<mrow> <mi>S</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <mn>0</mn> </mtd> <mtd> <msub> <mi>k</mi> <mn>12</mn> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mn>13</mn> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>1</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mn>21</mn> </msub> </mtd> <mtd> <mn>0</mn> </mtd> <mtd> <msub> <mi>k</mi> <mn>23</mn> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>2</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mn>31</mn> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mn>32</mn> </msub> </mtd> <mtd> <mn>0</mn> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>3</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>3</mn> </mrow> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <mn>0</mn> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>
Wherein, S is that the user follows relational matrix, and k is positive natural number, and N is to be made comments in the topic relevant information The quantity of the user of information, when S is that user replys relational matrix, kijThe comment delivered for i-th of user j-th of user of reply The number of information, when S is that user forwards relational matrix, kijThe comment information delivered for i-th of user j-th of user of forwarding Number, when S is user's adduction relationship matrix, kijThe number for the comment information that j-th of user delivers, i are quoted for i-th of user ∈ [1, N], j ∈ [1, N], i ≠ j；
As the kijDuring ＞ 0, i-th of user of line and j-th of user, letter of being made comments in the topic relevant information is built The topological diagram for following relation of each user of breath.
5. according to the method for claim 4, it is characterised in that it is described to follow relation based on described, calculate the topic The value of parameter preset, topic analysis report is generated, including：
The cluster coefficients of each user for information of being made comments in the topic relevant information are calculated according to below equation：
<mrow> <msub> <mi>C</mi> <mi>i</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mn>2</mn> <msub> <mi>E</mi> <mi>i</mi> </msub> </mrow> <mrow> <msub> <mi>l</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>l</mi> <mi>i</mi> </msub> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>;</mo> </mrow>
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in the topological diagram with i-th of user The neighbor user quantity of line, EiFor the wiring quantity between the neighbor user.
6. according to the method for claim 4, it is characterised in that it is described to follow relation based on described, calculate the topic The value of parameter preset, topic analysis report is generated, including：
The network density is calculated according to below equation：
<mrow> <mi>B</mi> <mo>=</mo> <mfrac> <mrow> <mn>2</mn> <mi>L</mi> </mrow> <mrow> <mi>N</mi> <mrow> <mo>(</mo> <mi>N</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>;</mo> </mrow>
Wherein, B is the network density, and L is the wiring quantity in the topological diagram.
7. according to the method for claim 4, it is characterised in that follow relation based on described described in network diameter, calculate institute The value of the parameter preset of topic is stated, generates topic analysis report, including：
The network diameter is calculated according to below equation：
D=maxdij；
Wherein, D represents the network diameter, and d represents the distance in the topological diagram between each user, dijRepresent in the topology Distance in figure between i-th of user and j-th of user, max are represented to the dijTake maximum.
8. according to the method for claim 4, it is characterised in that it is described to follow relation based on described, calculate the topic The value of parameter preset, topic analysis report is generated, including：
The value of each parameter preset of the topic is weighted superposition, obtains comprehensive analysis parameter；
Generation includes the topic analysis report of the comprehensive analysis parameter.
9. a kind of much-talked-about topic pusher, it is characterised in that described device includes：
Comment information acquiring unit, the comment information to topic delivered for obtaining user, and extract the key of the topic Word；
Relevant information determining unit, for calculating the comment information and the similarity of the keyword of extraction, similarity is more than The comment information of first preset value is as topic relevant information；
Follow relationship analysis unit, between each user for analyzing information of making comments in the topic relevant information with It is described to follow relation including at least one of following with relation：Reply relation, forwarding relation, adduction relationship；
Analysis report generation unit, for following relation based on described, the value of the parameter preset of the topic is calculated, generates topic Analysis report, the parameter preset include at least one of following：Network clustering coefficient, network density, network diameter, in the words The quantity of the user for information of being made comments in topic relevant information, wherein, the network clustering coefficient is by the related letter of the topic The cluster coefficients weighted average of each user for information of being made comments in breath and obtain, the network density is represented in the topic phase The tightness degree for following relation of each user for information of being made comments in information is closed, the network diameter is related in the topic The maximum of distance between two users of information of being made comments in information；
Much-talked-about topic push unit, for topic corresponding to the topic analysis report for meeting the first preparatory condition to be talked about as focus Topic, pushes the much-talked-about topic.
10. device according to claim 9, it is characterised in that the relevant information determining unit, including：
Word-dividing mode, for carrying out word segmentation processing to the comment information, obtain multiple words；
Computing module, for calculating the similarity of the multiple word and the keyword of extraction.
11. device according to claim 9, it is characterised in that it is described to follow relationship analysis unit, including：
Determining module is replied, the comment information for being delivered as user in the topic relevant information is the reply topic phase When closing the comment information that other users deliver in information, determine between two users to be reply relation；And/or
Determining module is forwarded, the comment information for being delivered as user in the topic relevant information is the forwarding topic phase When closing the comment information that other users deliver in information, determine between two users to be forwarding relation；And/or
Determining module is quoted, the comment information for being delivered as user in the topic relevant information is the reference topic phase When closing the comment information that other users deliver in information, determine between two users to be adduction relationship.
12. device according to claim 11, it is characterised in that the analysis report generation unit, including：
Follow relational matrix to establish module, for following relation based on described, user is established using below equation and follows relation square Battle array S：
<mrow> <mi>S</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <mn>0</mn> </mtd> <mtd> <msub> <mi>k</mi> <mn>12</mn> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mn>13</mn> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>1</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mn>21</mn> </msub> </mtd> <mtd> <mn>0</mn> </mtd> <mtd> <msub> <mi>k</mi> <mn>23</mn> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>2</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mn>31</mn> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mn>32</mn> </msub> </mtd> <mtd> <mn>0</mn> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mn>3</mn> <mi>N</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> <mtd> <mo>&amp;CenterDot;</mo> </mtd> </mtr> <mtr> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>k</mi> <mrow> <mi>N</mi> <mn>3</mn> </mrow> </msub> </mtd> <mtd> <mo>...</mo> </mtd> <mtd> <mn>0</mn> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>
Wherein, S is that the user follows relational matrix, and k is positive natural number, and N is to be made comments in the topic relevant information The quantity of the user of information, when S is that user replys relational matrix, kijThe comment delivered for i-th of user j-th of user of reply The number of information, when S is that user forwards relational matrix, kijThe comment information delivered for i-th of user j-th of user of forwarding Number, when S is user's adduction relationship matrix, kijThe number for the comment information that j-th of user delivers, i are quoted for i-th of user ∈ [1, N], j ∈ [1, N], i ≠ j；
Topological diagram builds module, for as the kijDuring ＞ 0, i-th of user of line and j-th of user, build in the topic The topological diagram for following relation of each user for information of being made comments in relevant information.
13. device according to claim 12, it is characterised in that the analysis report generation unit, in addition to：
Cluster coefficients computing module, for calculating the every of information that made comments in the topic relevant information according to below equation The cluster coefficients of individual user：
<mrow> <msub> <mi>C</mi> <mi>i</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mn>2</mn> <msub> <mi>E</mi> <mi>i</mi> </msub> </mrow> <mrow> <msub> <mi>l</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>l</mi> <mi>i</mi> </msub> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>;</mo> </mrow>
Wherein, wherein C is cluster coefficients, CiFor the cluster coefficients of i-th of user, liFor in the topological diagram with i-th of user The neighbor user quantity of line, EiFor the wiring quantity between the neighbor user.
14. device according to claim 12, it is characterised in that the analysis report generation unit, in addition to：
Network density computing module, for calculating the network density according to below equation：
<mrow> <mi>B</mi> <mo>=</mo> <mfrac> <mrow> <mn>2</mn> <mi>L</mi> </mrow> <mrow> <mi>N</mi> <mrow> <mo>(</mo> <mi>N</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>;</mo> </mrow>
Wherein, B is the network density, and L is the wiring quantity in the topological diagram.
15. device according to claim 12, it is characterised in that the analysis report generation unit, in addition to：
Network diameter computing module, for calculating the network diameter according to below equation：
D=maxdij；
Wherein, D represents the network diameter, and d represents the distance in the topological diagram between each user, dijRepresent in the topology Distance in figure between i-th of user and j-th of user, max are represented to the dijTake maximum.
16. device according to claim 12, it is characterised in that the analysis report generation unit, in addition to：
Comprehensive analysis parameter determination module, for the value of each parameter preset of the topic to be weighted into superposition, integrated Analytical parameters；
Topic analysis report generation module, the topic analysis report of the comprehensive analysis parameter is included for generating.
CN201610355128.9A 2016-05-25 2016-05-25 Hot topic pushing method and device Active CN107436877B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610355128.9A CN107436877B (en) 2016-05-25 2016-05-25 Hot topic pushing method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610355128.9A CN107436877B (en) 2016-05-25 2016-05-25 Hot topic pushing method and device

Publications (2)

Publication Number Publication Date
CN107436877A true CN107436877A (en) 2017-12-05
CN107436877B CN107436877B (en) 2021-03-30

Family

ID=60454094

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610355128.9A Active CN107436877B (en) 2016-05-25 2016-05-25 Hot topic pushing method and device

Country Status (1)

Country Link
CN (1) CN107436877B (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110096649A (en) * 2019-05-14 2019-08-06 武汉斗鱼网络科技有限公司 A kind of model extracting method, device, equipment and storage medium
CN110263237A (en) * 2019-05-31 2019-09-20 精硕科技（北京）股份有限公司 The acquisition methods and device of public sentiment data
CN111026981A (en) * 2019-12-05 2020-04-17 中国银行股份有限公司 Visual display method, device and equipment for hot topics
CN112948527A (en) * 2021-02-23 2021-06-11 云南大学 Improved TextRank keyword extraction method and device
CN113051484A (en) * 2019-12-27 2021-06-29 北京国双科技有限公司 Method and device for determining hot social information
CN114357278A (en) * 2020-09-28 2022-04-15 腾讯科技（深圳）有限公司 Topic recommendation method, device and equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070083491A1 (en) * 2004-05-27 2007-04-12 Silverbrook Research Pty Ltd Storage of key in non-volatile memory
CN101819585A (en) * 2010-03-29 2010-09-01 哈尔滨工程大学 Device and method for constructing forum event dissemination pattern
CN103218400A (en) * 2013-03-15 2013-07-24 北京工业大学 Method for dividing network community user groups based on link and text contents
CN103324666A (en) * 2013-05-14 2013-09-25 亿赞普(北京)科技有限公司 Topic tracing method and device based on micro-blog data
CN103793489A (en) * 2014-01-16 2014-05-14 西北工业大学 Method for discovering topics of communities in on-line social network
CN103793503A (en) * 2014-01-24 2014-05-14 北京理工大学 Opinion mining and classification method based on web texts

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070083491A1 (en) * 2004-05-27 2007-04-12 Silverbrook Research Pty Ltd Storage of key in non-volatile memory
CN101819585A (en) * 2010-03-29 2010-09-01 哈尔滨工程大学 Device and method for constructing forum event dissemination pattern
CN103218400A (en) * 2013-03-15 2013-07-24 北京工业大学 Method for dividing network community user groups based on link and text contents
CN103324666A (en) * 2013-05-14 2013-09-25 亿赞普(北京)科技有限公司 Topic tracing method and device based on micro-blog data
CN103793489A (en) * 2014-01-16 2014-05-14 西北工业大学 Method for discovering topics of communities in on-line social network
CN103793503A (en) * 2014-01-24 2014-05-14 北京理工大学 Opinion mining and classification method based on web texts

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110096649A (en) * 2019-05-14 2019-08-06 武汉斗鱼网络科技有限公司 A kind of model extracting method, device, equipment and storage medium
CN110096649B (en) * 2019-05-14 2021-07-30 武汉斗鱼网络科技有限公司 Post extraction method, device, equipment and storage medium
CN110263237A (en) * 2019-05-31 2019-09-20 精硕科技（北京）股份有限公司 The acquisition methods and device of public sentiment data
CN111026981A (en) * 2019-12-05 2020-04-17 中国银行股份有限公司 Visual display method, device and equipment for hot topics
CN111026981B (en) * 2019-12-05 2023-10-24 中国银行股份有限公司 Visual display method, device and equipment for hot topics
CN113051484A (en) * 2019-12-27 2021-06-29 北京国双科技有限公司 Method and device for determining hot social information
CN114357278A (en) * 2020-09-28 2022-04-15 腾讯科技（深圳）有限公司 Topic recommendation method, device and equipment
CN114357278B (en) * 2020-09-28 2024-03-19 腾讯科技（深圳）有限公司 Topic recommendation method, device and equipment
CN112948527A (en) * 2021-02-23 2021-06-11 云南大学 Improved TextRank keyword extraction method and device

Also Published As

Publication number Publication date
CN107436877B (en) 2021-03-30

Similar Documents

Publication Publication Date Title
Giannoulakis et al. Evaluating the descriptive power of Instagram hashtags
CN107436877A (en) Much-talked-about topic method for pushing and device
Jäschke et al. Tag recommendations in social bookmarking systems
CN103324665B (en) Hot spot information extraction method and device based on micro-blog
US9324112B2 (en) Ranking authors in social media systems
Agarwal et al. Modeling and data mining in blogosphere
CN107220386A (en) Information-pushing method and device
CN108897734A (en) User&#39;s portrait generation method, device, computer equipment and storage medium
US20120042020A1 (en) Micro-blog message filtering
CN106484764A (en) User&#39;s similarity calculating method based on crowd portrayal technology
CN103425763B (en) User based on SNS recommends method and device
Liu et al. Fractal and multifractal analyses of bipartite networks
US20180046628A1 (en) Ranking social media content
Ting Web mining techniques for on-line social networks analysis
JP5754854B2 (en) Contributor analysis apparatus, program and method for analyzing poster profile information
CN107526718A (en) Method and apparatus for generating text
CN105787066A (en) Digital content distribution system based on total analysis
Deng et al. User interest mining via tags and bidirectional interactions on Sina Weibo
Rani et al. A survey of tools for social network analysis
CN112116426A (en) Method and device for pushing article information
CN109190027A (en) Multi-source recommended method, terminal, server, computer equipment, readable medium
Jia et al. TT-graph: A new model for building social network graphs from texts with time series
CN111882224A (en) Method and device for classifying consumption scenes
Liu et al. Identifying website communities in mobile internet based on affinity measurement
Heravi et al. Tweet location detection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant