CN102075851B - Method and system for acquiring user preference in mobile network - Google Patents
Method and system for acquiring user preference in mobile network Download PDFInfo
- Publication number
- CN102075851B CN102075851B CN200910238504.6A CN200910238504A CN102075851B CN 102075851 B CN102075851 B CN 102075851B CN 200910238504 A CN200910238504 A CN 200910238504A CN 102075851 B CN102075851 B CN 102075851B
- Authority
- CN
- China
- Prior art keywords
- user
- context
- preference
- historical behavior
- user preference
- 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.)
- Expired - Fee Related
Links
Abstract
The invention discloses a method and system for acquiring user preference in a mobile network. The method comprises the following steps: establishing a user preference acquisition model based on context computation so as to divide available context information into a current context and a historical context; dividing the context computation into current context-aware computation and historical context computation; taking user historical behavior and context thereof as main data sources, and accurately mining the user preference and change thereof step by step by utilizing methods such as collaborative filtering, historical context computation and the like; designing a user preference acquisition system in the mobile network according to the user preference acquisition model by combination with the actual situation of the mobile network so as to inspect the fitting effect of a user preference extraction model in the actual scene of the mobile network and further verify the availability and advanced property of the user preference extraction model; and finally achieving the purposes of extracting group user preference and individual user preference based on different objects from massive information.
Description
Technical field
The present invention relates to the business support technology of moving communicating field, particularly relate to acquisition methods and the system of user preference in a kind of mobile network.
Background technology
Along with 3G (3G (Third Generation) Moblie), 4G (forth generation mobile communication) network architecture is towards syncretization, complanation future development, mobile communications network is on the basis of merging gradually with computer network, conventional internet information service is extended, for user provides more colourful mobile network service and be not only simple communication service, as: traditional voice and data service, mobile search is served, Location based service, wirelessly browse class business, mobile stream medium service, mobile electronic payment is served, mobile download service, mobile network game etc., the mobile network service that various Mobile Network Operator and service provider provide is in content, price, QoS (Quality of Service, service quality) etc. also there is larger difference.Because Intelligent mobile equipment is day by day universal, the acquisition of information resources and pushing can occur in " any time, any place, any mode ", and mobile subscriber will be in complicated communication network environment and abundant service provides environment.But, when the interface display, terminal processes, input and output etc. of mobile device are limited in one's ability, the mobile network service data of magnanimity, dynamic change often for user brings very heavy information burden, thus will cause " mobile message overload " problem.
Thus, mobile network service select permeability, namely how accurately provide its real interested mobile network service type and content for mobile subscriber, just seem particularly important and urgent, risen to the key issue being badly in need of solving in mobile network service field thus more and more paid close attention to.Such as, some research institutions start the Individuation research paying close attention to next generation network service field both at home and abroad, and the enterprises such as China Mobile, Google company also start the development launching mobile personalized service product.If this problem can solve smoothly, the mobile network service that not only can improve user is experienced, and also can maintain customer group for Mobile Network Operator and service provider, the good service of realization " people-oriented " plays great facilitation.
Mobile Network Operator and service provider only, after abundant, accurate understanding user is to the demand of various service and change thereof, just can provide the service meeting its demand and content; Therefore, building accurate user preference and obtain model to obtain the individual demand of user, is the most key technology.But the complicacy due to mobile communications network: on the one hand, mobile network service is in the environment of height dynamic change, relate to the development of infrastructure network, operation maintenance management, service logic design, content production and distribution, terminal device etc., its application is extensive, closely related with the feature of each application industry, and the propelling movement of mobile network service and content thereof can occur in " any time, any place, by any way ", makes the complexity of research work very high; On the other hand, mobile subscriber has different Demographics backgrounds and context environmental, progressively ripe to the degree of awareness of various service, be also just not quite similar and increasingly complex to the demand of serving and preference, this turn increases the difficulty of research work undoubtedly.Therefore, in mobile network user preference acquiring technology be one be worth research and challenging work.
At the end of last century, along with the fast development of Internet technology and the appearance of user's " information overload " and " information puzzle " problem, people propose the concept of " personalized service " providing Differentiated Services for different user.The personalized service research in conventional internet field obtains the interest level of user to project to build user preferences modeling, mainly based on the two-dimensional space of user-project (user-item), the technology such as collaborative filtering, content-based filtering, composite filtering that depend on realize.Although the Individuation research in conventional internet field achieves many achievements, great majority are for desktop soft and hardware system " information overload " problem.
At present, along with the develop rapidly of mobile communications network and the individual demand of user to Information Mobile Service more and more higher, domestic and international researchist starts to study in mobile personalized service, and mobile subscriber's preferential learning also starts to be paid close attention to as the gordian technique of mobile personalized service.Such as, document " G.Lee; S.Bauer; P.Faratin; J.Wroclawski.Learning User Preferences for Wireless ServicesProvisioning.2004:P480-487 ", for dynamic radio service selection problem, proposes a kind of method utilizing nitrification enhancement and Markov model study user preference; Document " Sheng; Q.; B.Benatallah; and Z.Maamar; User-centric services provisioning in wireless environments.Communication of the ACM; Nov, 2008.51 (14): P130-135 " describes a kind of mechanism providing service content in wireless network environment towards different user; Document " H.J.Lee; S.J.Park; MONERS:A News Recommender for the Mobile Web.Expert Systems with Applications; 2007; 32 (1): P143-150 " then proposes a kind of news commending system towards mobile Internet, recommend news by theme of news self importance and ageing, user preference changes, user decides for the preference of news generic.Eighties of last century beginning of the nineties, the concept that Weiser proposes " general fit calculation ", the context-aware computing as one of the sub-field of its core starts fully to be paid close attention to.Here, the research contents of described context-aware computing mainly comprises: context obtains, context modeling and expression, contextual effective utilization, how to build the system framework etc. supporting context-aware.Its target system is found automatically and provides service and computational resource for user with taking the photograph before utilizing the contextual information such as position, surrounding environment, thus reduce man-machine interaction, improves Consumer's Experience.It is occur a kind of new technology trends recent years that described context-aware computing combines with information network.In next generation network (NGN) environment merged, context-aware theory is used for Physical layer or network layer (as wireless sensor network), then relatively less in the application of operation layer (particularly towards the individual business logic of mobile subscriber), such as: document " Cheng Bo, MengXiangwu, Chen Junliang.An Adaptive User Requirements Elicitation Framework.IEEE Computer Society Washington, DC, USA, 2007:P501-502 " consider the effect that context-aware computing extracts user's request, framework is extracted in the user's request proposing a kind of ontology-driven, document " Liu Dong, Meng Xiangwu, Chen Junliang.A Framework for Context-AwareService Recommendation.Advanced Communication Technology, ICACT 2008:P2131-2134 " proposes a kind of context-aware service and recommends framework.
But in mobile network service research field, due to user to the demand of mobile network service and content thereof or preference usually relevant to various contextual information, and it is more complicated than the context environmental in conventional internet field, therefore, the user preference that legacy user's preference pattern is not exclusively applicable under mobile network service environment obtains, and traditional recommendation system framework is also not suitable for the modeling of mobile network service preference pattern.In a word, in above-mentioned research, context-aware computing is used for greatly recommending production process, and less for user preference acquisition process, also lacks and to obtain the user preference calculated based on context in mobile network and the further investigation of services selection model.
Visible, although comparatively deep for the research of the technology such as personalized service, context-aware computing, user preference acquisition both at home and abroad, comparatively horn of plenty is studied also for UNE service of future generation, service oriented computing scheduling theory, but be still in the exploratory stage for the research of user preference acquiring technology in mobile network, urgently further investigate further.
Summary of the invention
In view of this, fundamental purpose of the present invention is the acquisition methods and the system that provide user preference in a kind of mobile network, user preference is obtained in conjunction with context-aware computing theory according to the use habit of user, provide personalized service to user, experience with the network service improving mobile subscriber, meanwhile, technological means can be adopted to filter a large amount of redundant informations, make full use of Internet resources, improve the service quality of Virtual network operator and service provider and cut operating costs further.
For achieving the above object, technical scheme of the present invention is achieved in that
An acquisition system for user preference in mobile network, described acquisition system comprises user's historical behavior and context generates subsystem (21), data storage and management subsystem (22), data mining subsystem (23) and user preference extraction subsystem (24); Wherein:
User's historical behavior and context generate subsystem (21), in order to complete mobile subscriber's historical behavior and the contextual data genaration of user's historical behavior;
Data storage and management subsystem (22), in order to the store and management of completing user historical behavior data, user's historical behavior contextual information, user preference information;
Data mining subsystem (23), calculates in order to completing user cluster and user's historical behavior context;
User preference extracts subsystem (24), for the result of calculation according to described data mining subsystem (23), extract group user preference and individual consumer's preference information, and output to described data storage and management subsystem (22).
Wherein, described acquisition system comprises further:
User preference self-adaptation subsystem (25), in order to the change/collision detection of completing user preference, user preference correction, and is kept at the result of testing result or correction in data storage and management subsystem (22).
Described user's historical behavior and context generate subsystem (21) and comprise further: user's historical behavior generation module (211) and user's historical behavior context generation module (212); Wherein,
Described user's historical behavior generation module (211), for realizing the data genaration function of user's historical behavior, its Output rusults is the data source of group user preference extraction;
Described user's historical behavior context generation module (212), for realizing the contextual data genaration function of user's historical behavior, its Output rusults is the data source that user's historical behavior context calculates.
Described data mining subsystem (23), comprises user clustering module (231) and user's historical behavior context computing module (232) further; Wherein,
Described user clustering module (231), based on the use amount of user to Information Mobile Service, by using clustering algorithm, all users are divided in multiple different cluster, make the user's similarity in same cluster higher, user's similarity in different cluster is lower, and after cluster analysis terminates, each user has a cluster labelled notation;
Described user's historical behavior context computing module (232), for calculating user's historical behavior context, in the hope of going out individual consumer to contextual interest-degree in a certain respect.
Described user preference extracts subsystem (24), comprises group user preference extraction module (241) and individual consumer's preference extraction module (242) further; Wherein,
Described group user preference extraction module (241), for calculating group user cluster result and group user historical behavior, exports group user preference information;
Described individual consumer's preference extraction module (242), for carrying out fusion calculation to group user preference and user's historical behavior context result of calculation, exports individual consumer's preference information.
An acquisition methods for user preference in mobile network, the method comprises:
A, utilize user's historical behavior and context to generate subsystem (21) to generate user's historical behavior data and user's historical behavior context data, and described data are kept in data storage and management subsystem (22);
B, from described data storage and management subsystem (22), obtain user's historical behavior and context data by data mining subsystem (23), and carry out user clustering respectively and user behavior historical context calculates;
C, extract subsystem (24) by user preference again and extract group user preference according to user clustering result, and then group user preference and user's historical behavior context result of calculation are carried out fusion calculation, to extract individual consumer's preference, and user preference information is stored in described data storage and management subsystem (22).
Comprise further after described step C:
User preference self-adaptation subsystem (25) is according to user feedback, detect the change of user's historical behavior and context data thereof, or detect and extracted conflicting of user preference and real user demand, then adaptive correction is carried out to the user preference extracted, and user will be presented to through revised Output rusults by mobile network service and content thereof.
Based on individual consumer's interest-degree computing method that context calculates in mobile network described in claim 1, the method comprises:
The case data information of A, extraction user historical behavior context computing module, reads user's historical behavior context database, reads processed contextual information;
B, according to read user's historical behavior contextual information inquiry case database mate, check whether case exists, if case exist, then perform step C; Otherwise, perform step D;
C, the corresponding case in case library to be modified, then perform step e;
D, by the amendment made case or the new case created stored in case database, then perform step e;
E, to judge whether that contextual information has read complete, if do not read, then return and perform steps A and continue to read contextual information, until described contextual information reads complete, case library creates complete; Otherwise, perform step F;
F, end case extract flow process, start context calculation process, read case information in case library, by calculating the precondition value in Bayesian network, and according to the conditional value that previous step has obtained, calculate the value of the Bayesian network of each subnet of separation, draw the daily behavior custom of user, then by the value of each subnet that draws stored in database;
G, calculate the service probability of use value of user according to the value of each precondition and each subnet, the service probability of use value of clustering algorithm to user is used to carry out cluster analysis, so that probable value is divided into different grades, to extract individual consumer's interest-degree, and by extracted individual consumer's interest-degree stored in individual consumer's interest-degree database.
Merge a user preference extracting method for collaborative filtering and context calculating in mobile network described in claim 1, the method comprises:
A, acquisition individual consumer's interest-degree and group of subscribers preference, travel through each user under often kind of context environmental to the interest-degree that every class is served; B, judge whether user has traveled through, if do not traveled through, then perform step c; Otherwise, terminate this ergodic process; C, judge whether the traversal of serving in certain class for certain individual consumer completes, if complete, returns step b; Otherwise, perform steps d; D, judge for described individual consumer certain class service under context traversal whether complete, if complete, return step c; Otherwise, perform step e; Whether e, the interest-degree judging described individual consumer are zero, if so, then perform step f; Otherwise, perform step I; F, travel through this individual consumer place group every other user under this kind of context environmental to the interest-degree of such service, and obtain this user place group identification, then perform step g; G, calculate described every other user effective mean value of interest-degree to such service under this kind of context environmental; H, judge whether described effective mean value is zero, if so, then performs step I; Otherwise, perform step j; I, individual consumer's preference value are the preference value that this user place group serves such, then perform step l; J, individual consumer's preference value are effective mean value; Then step l is performed; K, individual consumer's preference value are the interest-degree of this individual consumer, then perform step l; L, individual consumer's preference is stored in corresponding database, then returns execution steps d.
The acquisition methods of user preference and system in mobile network provided by the present invention, have the following advantages:
The present invention is a kind of based on user preference acquisition model in the mobile network of context calculating by proposing, and design corresponding prototype system to realize the validity obtaining model with this user preference, wherein, by using user's historical behavior context generation module, context environmental when recording user historical behavior occurs residing for user, and these historical context are calculated, thus solve the user preference calculated based on context, will the degree of accuracy improving personalized service be conducive to; The user interest degree computing method calculated based on context are adopted in user's historical behavior context computing module, reasoning is carried out to historical context and utilizes bayesian theory to calculate user's historical behavior context, solve various context probability when user's historical behavior occurs, and the individual consumer's interest-degree finally utilizing the form of various dimensions matrix to describe, the precision having enriched user preference describes; The user preference extracting method adopting and merge collaborative filtering and context calculating is extracted in subsystem at user preference, the service preferences of the individual consumer's interest-degree calculated based on context and individual consumer place group is carried out fusion calculation, thus predict and extract different user to the preference of different mobile network service under different context environmental, the user preference information described with various dimensions matrix form needed for final generation.In addition, the user preference adaptive approach calculated based on context is implemented in user preference self-adaptation subsystem, by detecting the change of user preference and conflict, and the result of user preference information being revised, further increasing the degree of accuracy extracting user preference information.
Accompanying drawing explanation
Fig. 1 is that the user preference calculated based on context in the embodiment of the present invention obtains model schematic;
Fig. 2 is that in mobile network of the present invention, user preference obtains system architecture schematic diagram;
Fig. 2 A is user's historical behavior generation module illustrative view of functional configuration that in Fig. 2 of the present invention, user's historical behavior and context generate subsystem;
Fig. 2 B is user's historical behavior context generation module illustrative view of functional configuration that in Fig. 2 of the present invention, user's historical behavior and context generate subsystem;
Fig. 2 C is the user clustering functions of modules structural representation of data mining subsystem in Fig. 2 of the present invention;
Fig. 2 D is user's historical behavior context computing module illustrative view of functional configuration of data mining subsystem in Fig. 2 of the present invention;
Fig. 2 E is the group user preference extraction functions of modules structural representation that in Fig. 2 of the present invention, user preference extracts subsystem;
Fig. 2 F is individual consumer's preference extraction functions of modules structural representation that in Fig. 2 of the present invention, user preference extracts subsystem;
Fig. 2 G is the illustrative view of functional configuration of user preference self-adaptation subsystem in Fig. 2 of the present invention;
Fig. 3 is the physical arrangement schematic diagram that in the embodiment of the present invention, mobile network user preference obtains system;
Fig. 3 A is the base conditioning schematic flow sheet that in mobile network of the present invention, user preference obtains system;
Fig. 4 is historical behavior context generative process schematic diagram in prototype system of the present invention;
Fig. 5 is the schematic flow sheet based on individual consumer's interest-degree computing method of context calculating in prototype system of the present invention;
Fig. 6 is the user preference extracting method schematic flow sheet that fusion collaborative filtering in prototype system of the present invention and context calculate.
Embodiment
Below in conjunction with accompanying drawing and embodiments of the invention, method of the present invention is described in further detail.
Core concept of the present invention is: obtain model by setting up a user preference calculated based on context, on the basis of traditional personalization service system, take into full account the importance of the contextual information of complicated dynamic change in mobile network user preference obtains, available contextual information is divided into current context and historical context, and therefore contextual calculating be divided into current context perception calculating and historical context to calculate, then using user's historical behavior and context thereof as main data source, by method Query refinement digging user preference and changes thereof such as collaborative filtering and historical context calculating, then model is obtained according to described user preference, in conjunction with the actual conditions of mobile phone users, Mobile Network Operator and Internet Service Provider, design user preference in a mobile network and obtain system, to check described user preference extraction model to the fitting effect in mobile network's actual scene, thus verify validity and the advance of described user preference extraction model.Described prototype system essence is an independently system, its function is cooperatively interacted by some subsystems and module, is jointly completed, its function is corresponding with the function that described model should complete, and prototype system mainly comprises: user's historical behavior and context generate subsystem, data storage and management subsystem, data mining subsystem, user preference extraction subsystem, user preference self-adaptation subsystem and user interface.Generate subsystem, data storage and management subsystem, data mining subsystem, user preference self-adaptation subsystem and user preference by utilizing user's historical behavior of described prototype system and context and extract subsystem, finally realize extracting the object of group user preference for different object and individual consumer's preference from magnanimity information.
Fig. 1 is that the user preference calculated based on context in the embodiment of the present invention obtains model schematic, as shown in Figure 1, this user preference obtains model and belongs to data driven type model, it can be divided into data collection layer from low to high from functional plane, user preference securing layer and user preference adaptation layer, this model uses user preference extractive technique and user preference adaptive technique, the object of its research, i.e. data message, mainly from the service provider of mobile network, described data message comprises: user's historical behavior information, user's historical behavior contextual information, user's demographic information, mobile network service information, contextual information etc., also three classes can be divided into by the source of data: i.e. user profile, resource information and contextual information.
The mathematical description that user preference based on context calculating obtains model is tentatively: M={U, I, C, P}, U × I × C → P.Wherein: U representative of consumer information, I represents object resource information, and C represents contextual information, P representative of consumer preference.Model basis is user's historical behavior and user's historical behavior context mainly, is obtained by " data collection layer " of model bottom.Wherein, user's historical behavior is for describing the service condition of user to object resource ("current" model is object with mobile network service); User's historical behavior context is for describing context condition residing when user uses object resource.
Extraction due to user preference is the process of a Query refinement, user's " preference extraction layer " in Fig. 1 institute representation model first will extract comparatively coarse group user preference by calculating user's historical behavior, then individual consumer's interest-degree is extracted by calculating user's historical behavior context, last again by two kinds of data fusion calculating, extract comparatively accurate individual consumer's preference.
Here, user preference self-adaptation, refer to detect user preference change or extract conflicting of user preference and real user demand, and carry out the technology revised.User preference adaptive technique is realized by described " user preference adaptation layer ", and it mainly comprises two necessary processes: the first, user preference change/collision detection; The second, user preference correction.
The structure obtaining model due to user preference of the present invention realizes user preference in described mobile network to obtain the basis of system, therefore need to obtain to described user preference the theory or technology that use in model, as: the calculating of body, collaborative filtering, context, matrix theory and Markovian process etc. are done one and are simply introduced:
Described body, the clear and definite Formal Specification explanation of shared ideas model.Based on body, modeling is carried out to user, object resource, context, user preference etc., be conducive to carrying out structuring, formalized description from semantic level to above-mentioned every field data, be convenient to infer the carrying out of the complex relationship of every field data.
Described collaborative filtering, collaborative filtering method utilizes the similarity (or the similarity between project) between user, and the evaluation and recommendations based on other people provide decision support for user.The process that group user preference is extracted is mainly used in the present invention.
Described context calculates, and refers to that system can find and effectively utilize contextual information (as customer location, time, environmental parameter, contiguous equipment and personnel, User Activity etc.) to carry out a kind of computation schema calculated.Context is divided into two kinds by the model proposed in invention, i.e. historical context, current context; Thus context calculates and is also divided into two parts, i.e. historical context calculating, current context perception calculate.When obtaining user preference in the present invention, historical context when considering that user's historical behavior occurs residing for user; When producing recommendation, consider the context that user is current.
Described matrix theory, user preference obtains model and investigates user preference from multidimensional angles such as user, service, contexts, thus use multi-dimensional matrix theory to calculate each denapon element affecting user preference, and the user preference got is stored in multi-dimensional matrix.
Described Markov process, refers at a time event, and the stochastic process just had an impact in limited period of time, belongs to theory of random processes.Markov process theory is applied in user preference adaptive polo placement process, effectively can detects and compute user preferences change, thus carry out user preference correction.
Fig. 2 is that in mobile network of the present invention, user preference obtains system architecture schematic diagram, as shown in Figure 2, described prototype system generates subsystem 21, data storage and management subsystem 22, data mining subsystem 23 primarily of user's historical behavior and context, user preference extracts subsystem 24 and user preference self-adaptation subsystem 25 is formed.Wherein,
User's historical behavior and context generate subsystem 21, in order to complete mobile subscriber's historical behavior and the contextual data genaration of user's historical behavior.
Here, described user's historical behavior and context generation subsystem 21 also comprise: user's historical behavior generation module 211, user's historical behavior context generation module 212.
Described user's historical behavior generation module 211, for realizing the data genaration function of user's historical behavior, its Output rusults can as the data source of group user preference extraction.
Described user's historical behavior context generation module 212, for realizing the contextual data genaration function of user's historical behavior, the data source that its Output rusults can calculate as user's historical behavior context.
It should be noted that, those are only selected to affect the most deep context type of user preference in the present invention, as time, position, use equipment and User Activity state etc., but this does not limit the described user preference calculated based on context and obtains the context that above-mentioned several types only supported by model, described model does not also rely on concrete context type, can also further expand, as increased the social environment etc. of weather, noisy environment, light, user.
Data storage and management subsystem 22, in order to the store and management of completing user historical behavior data, user's historical behavior contextual information, user preference information.
Here, described data storage and management subsystem 22, comprises user preference storage and management module 221, user's historical behavior storage and management module 222 and user's historical behavior context storage and management module 223 further.Wherein: described user preference storage and management module 221, for store and management user preference information.Described user's historical behavior storage and management module 222, for store and management user historical behavior information.Described user's historical behavior context storage and management module 223, for store and management user historical behavior contextual information.
Described data storage and management subsystem 22, the tables of data relevant by user message table, user's historical behavior table, user's historical behavior contextual information table, group user preference table, individual consumer's preference table and user interface manages Various types of data information.Data storage and management subsystem 22 mainly manages data message user message table, user's historical behavior table, user's historical behavior contextual information table, group user preference table, individual consumer's preference table and the data structure that is associated thereof.
Data mining subsystem 23, calculates in order to completing user cluster and user's historical behavior context.
Here, described data mining subsystem 23, comprises user clustering module 231 and user's historical behavior context computing module 232 further; Wherein,
Described user clustering module 231, based on the use amount of user to Information Mobile Service, by using K-Means clustering algorithm, all users are divided in the individual different cluster of K, make the user's similarity in same cluster higher, user's similarity in different cluster is lower, and after cluster analysis terminates, each user has a cluster labelled notation.
User's historical behavior context computing module 232, for calculating user's historical behavior context, in the hope of going out individual consumer to interest-degree in a certain respect.
User preference extracts subsystem 24, for the result of calculation according to described data mining subsystem 23, extracts group user preference and individual consumer's preference information and export from described data storage and management subsystem 22; Also for the detection/correction result according to described user preference self-adaptation subsystem 25 and extract in conjunction with the result of calculation of described data mining subsystem 23 and export revised group user preference and individual consumer's preference information.
Here, described user preference extracts subsystem 24, comprises group user preference extraction module 241 and individual consumer's preference extraction module 242 further; Wherein,
Described group user preference extraction module 241, for calculating group user cluster result and group user historical behavior, exports group user preference information.
Described individual consumer's preference extraction module 242, for carrying out fusion calculation to group user preference and user's historical behavior context result of calculation, exports individual consumer's preference information.
User preference self-adaptation subsystem 25, in order to the change/collision detection of completing user preference, user preference correction, and is kept at the result of testing result or correction in data storage and management subsystem 22.
Here, user preference self-adaptation subsystem 25, comprises user preference change/collision detection module 251 and user preference correcting module 252 further.Wherein:
Described user preference change/collision detection module 251, for catching the change of user preference in time according to the change threshold set in advance or change detection algorithm; Or for detecting when the user preference that multi-source channel extracts merges, user preference may be occurred inconsistent and produce the phenomenon of conflicting, or detect extract and exist inconsistent between user preference and the preference of the explicit setting of user and produce the phenomenon of conflicting, and described user preference change/conflict situations is passed to user preference correcting module 252.
Described user preference correcting module 252, for the debugging functions of completing user preference, namely for the different objective circumstances that user preference conflict and user preference change, setting user preference correction algorithm, to reach, user preference information is revised, make the effect of obtained user preference information precision further.
In addition, outside prototype system, also comprise the user interface of the information interaction of a responsible user and described prototype system.Here, described user comprises mobile phone users and system management maintenance customer.
Fig. 2 A is user's historical behavior generation module illustrative view of functional configuration that in Fig. 2 of the present invention, user's historical behavior and context generate subsystem, and as shown in Figure 2 A, user's historical behavior generation module 211, comprises following submodule further:
Read external information document submodule 2111, resolve for information document to external world, and by analysis result stored in appointment array for other submodules.
Initialising subscriber information submodule 2112, for generating user ID, by it stored in database, and gives each attribute field assignment of this user.
Initialising subscriber behavior table submodule 2113, for carrying out initialising subscriber historical behavior table according to the user ID (major key) in user message table, described user's historical behavior table comprises major key (user ID) and each volume of services of each bar record in user behavior table.
Generate user behavior data submodule 2114, for reading analysis result array in external information document submodule 2111 as input, and by the Data Update that generates to the respective field in database user behavior table.
User's historical behavior data submodule 2115, for realizing the increase and decrease of user's historical behavior data, the basis as historical data in the month before increasing and decreasing, and is saved in database.
DB Backup submodule 2116, for fulfillment database information backup, so that check the data that early stage produces.
Fig. 2 B is user's historical behavior context generation module illustrative view of functional configuration that in Fig. 2 of the present invention, user's historical behavior and context generate subsystem, and as shown in Figure 2 B, user's historical behavior context generation module 212, comprises following submodule further:
User's historical behavior context generates father's agent sub-module 2121, for starting parent reason, promoter is acted on behalf of, receive the control message of superior agency, and from database, obtain user's volume of services data, inlet module is generated as whole user's historical behavior contextual information, this submodule controls operating procedure and the process of other module, the agency of this submodule controls the operation of self each module by the control information receiving higher level's submodule and user's historical behavior generation module, when after the control message receiving higher level's submodule, this submodule agency starts to start other and respectively acts on behalf of, and pass through the operation rhythm of each proxy module of user profile data flow con-trol.
User profile generates agent sub-module 2122, for generating the user profile of user, the user profile of generation is stored in user message table, by this agent sub-module, the information setting that can be user according to different situations meets oneself mechanics, after this agent sub-module runs, can be that user generates the user profile meeting certain rule according to the rule of user's setting.
Facility information generates agent sub-module 2123, for being each service creation subscriber equipment, each use is made to record corresponding a kind of subscriber equipment, and by the contextual information of generation stored in user context information table, with above-mentioned module, this submodule also can arrange different rules, the user of different characteristic is allowed to use different equipment, owing to being separate between each agent sub-module, therefore when arranging different rules, the logic of the current agent that other agency seldom can affect, and this change acting on behalf of internal logic can not have influence on the realization of other Agent logic completely.
Temporal information generates agent sub-module 2124, for the time context for each service creation user, make the time that each record correspondence one is different, with agent sub-module described above, this agent sub-module also can generate oneself logic according to the rule of self-demand of setting, the amendment of the logic of this agency does not affect other agency, bears results stored in the historical behavior context database table of relative users.
Positional information generates agent sub-module 2125: for the context for each service generation user locations, makes each use the corresponding user position of record; The formation logic of this agency, can completely according to different users, and different identity informations, produces different positional informations, and the contextual location information generated like this, by very close to the contextual information in reality, can provide better foundation for work from now on.
Action message generates agent sub-module 2126: be the action message context of each service creation user historical behavior, makes each use all corresponding a kind of User Status of record.With above-mentioned agent sub-module in like manner, according to different user profile, generate different user activity information, the rule that action message produces is arranged in described agent sub-module inside, different rules is set according to different environments for use, thus final generation meets the contextual information record of user's request.
Volume of services assignment agent submodule 2127: the volume of services of user's historical behavior is distributed, in the per a period of time allowing the volume of services of various service well be assigned in 30 days of one month, the distribution of volume of services can make random by totalizing method in all skies, also can according to must use habit distribution services amount, service logic in a word in agency can change at any time, different environment can be adapted to by the change of different service logics, meet various demand.
Fig. 2 C is the user clustering functions of modules structural representation of data mining subsystem in Fig. 2 of the present invention, and as shown in Figure 2 C, user clustering module 231 comprises following submodule further:
Extract user profile submodule 2311, for extract from the database depositing user's historical behavior information No. ID of user with user's historical behavior information, the use amount of namely serving.
Data-mapping submodule 2312, for being mapped on same interval [a, b] by all data according to certain reflection method, object is to reduce influencing each other between multidimensional data attribute, avoids large number to flood decimal, thus makes Clustering Effect more desirable.
Find cluster centre submodule 2313, for generation of K mutually different random number determination initial cluster center, and select initial cluster center.
Calculate cluster centre submodule 2314, for calculating the cluster centre of each cluster; Calculate the method for cluster centre be the mean value of all data in each cluster as cluster centre, specific practice is: multidimensional data calculates the mean value often tieed up.
Clustering submodule 2315, for by all Data Placement in cluster nearest with it, division methods is the Euclidean distance calculating data and each cluster centre.
Algorithm convergence judges submodule 2316, for being judged by the condition of convergence whether clustering terminates, if meet the condition of convergence, then terminates clustering algorithm; Otherwise, then clustering is proceeded, until meet the condition of convergence.
User clustering number arranges submodule 2317, for arranging a cluster number for each user in database.
Fig. 2 D is user's historical behavior context computing module illustrative view of functional configuration of data mining subsystem in Fig. 2 of the present invention, and as shown in Figure 2 D, user's historical behavior context computing module 232, comprises following submodule further:
The < time | position | movable, Information Mobile Service > case extracts submodule 2321, this module uses recorded information by obtaining the context generated, the < time is set up by extraction time wherein, position, activity, information on services | position | movable, Information Mobile Service > case library, in this case leaching process, read the context record information of generation one by one, set up case, the case repeated is removed, is only retained single case.Corresponding case library is had so that next step solves to each user.
The < time, position > daily habits case extracts submodule 2322, this module has generated context use recorded information by obtaining, and sets up the < time, position > case library by extraction time, positional information.In this case leaching process, after determining the demand to case type, read the context record information of generation one by one, set up case, the case repeated is removed, only retains single case.Corresponding case library is had so that next step solves to each user.
< position, movable > daily habits case extracts submodule 2323, this module has generated context use recorded information by obtaining, < position is set up, the case library of movable > by extracting position, activity contexts information.In this case leaching process, after determining the demand to case type, read the context record information of generation one by one, set up case, the case repeated is removed, only retains single case.Corresponding case library is had so that next step solves to each user.
The < time | position | movable, Information Mobile Service > interest-degree calculating sub module 2324, according to the < time | position | movable, the case that Information Mobile Service > interest-degree calculating sub module is extracted, calculate the interest-degree of user for corresponding case, because value that case in leaching process is corresponding is different (use amount as Information Mobile Service is different), thus different case finally to calculate gained interest level also different.By the calculating using Bayesian formula the calculating of posterior probability can be changed into prior probability, calculating of interest-degree is made well to solve and realize.Directed acyclic graph and probability theory organically combine by Bayesian network, cause-effect relationship digraph is showed intuitively, and the statistical value of historical record is all discrete data, Bayesian formula is made to have good effect being applied in the statistical computation of mobile field historical record.
The < time, position > daily habits interest-degree calculating sub module 2325, according to the < time, the case that position > daily habits interest-degree calculating sub module is extracted, calculate the interest-degree of user for corresponding case, the time value corresponding due to case in leaching process is different, thus different case finally to calculate gained interest level also different.By Bayesian formula, the calculating of posterior probability is changed into the calculating of prior probability, make calculating of interest-degree well solve and realize.
< position, movable > daily habits interest-degree calculating sub module 2326, according to < position, the case that movable > daily habits interest-degree calculating sub module is extracted, calculate the interest-degree of user for corresponding case, because the value that case in leaching process is corresponding is different, thus different case finally to calculate gained interest level also different.The calculating of posterior probability can be changed into the calculating of prior probability again by Bayesian formula, corresponding interest level can be calculated simply efficiently.
Fig. 2 E is the group user preference extraction functions of modules structural representation that in Fig. 2 of the present invention, user preference extracts subsystem, and as shown in Figure 2 E, group user preference extraction module 241, comprises following submodule further:
Group history behavioural matrix constructor module 2411, for completing the function of group history behavioural matrix structure, namely with user clustering result for input, calculate the average use amount of all users to all kinds of mobile network service of each group, structure group historical behavior matrix.
Service cluster submodule 2412, for completing the function of service cluster, namely with group history behavioural matrix for input, be object to be clustered with the historical behavior data of each group, according to clustering algorithm, each group's service condition cluster of every class being served is 5 (or 3) grades, that is: 5-very preference, 4-is preference comparatively, 3-general preference, and 2-is preference more not, 1-very not preference (or, the senior preference of 3-, 2-general preference, 1-is preference more not).
Group user preference calculating sub module 2413, for completing group user preference computing function, namely with group history behavioural matrix and service cluster result for input, according to service cluster result, group history behavioural matrix is mapped as group user preference matrix, wherein, described matrix adopts the two dimensional form of " group-service ", row vector is called with group, with all kinds of service for column vector, group user preference matrix is stored in the group user preference storage and management module in data storage and management subsystem.
Group user preference structure descriptor module 2414, for completing the structural description function of group user preference, namely with group user preference extraction result for input, describe document according to the XML Schema designed to the semi-structured group user preference generating XML format, and document is stored into group user preference storage and management module.
Fig. 2 F is individual consumer's preference extraction functions of modules structural representation that in Fig. 2 of the present invention, user preference extracts subsystem, and as shown in Figure 2 F, individual consumer's preference extraction module 242, comprises following submodule further:
Individual consumer's interest-degree zero value detection submodule 2421, for completing individual consumer's interest-degree zero value detection function, whether the interest-degree that namely in individual consumer's interest-degree database, certain user serves certain class under certain specific context condition is zero; If be zero, then perform individual consumer's preference calculating sub module function; Otherwise, perform effective average interest degree calculating sub module.
Effective average interest degree calculating sub module 2422, for completing effective average interest degree computing function, namely effective average interest degree calculating sub module needs other users finding group belonging to individual consumer from group user preference matrix database, and in individual consumer's interest-degree database, travel through these users under certain specific context condition to the nonzero value interest-degree that certain class is served, and calculate their mean value, export as effective average interest degree.
Effective average interest degree zero value detection submodule 2423, mainly completes effective average interest degree zero value detection function, namely detects whether effective average interest degree that effective average interest degree calculating sub module exports is zero.If be zero, then perform individual consumer's preference calculating sub module function, and do not access group's user preference database; Otherwise, perform individual consumer's preference computing module function, and access group user preference database.
Individual consumer's preference calculating sub module 2424, for completing individual consumer's preference computing function, be specially: according to result of calculation and the Rule of judgment of above-mentioned module, under different conditions, the value of individual consumer's preference is set to respectively individual consumer's interest-degree, effectively average interest degree or individual place group user preference, thus dope this user preference value to inhomogeneity service under different context environmentals, and Output rusults is stored in individual consumer's preference database (individual consumer's preference storage and management module).
Individual consumer's preference structure descriptor module 2425, for completing the structural description function of individual consumer's preference, namely with individual consumer's preference extraction result for input, describe document according to the XML Schema designed to the semi-structured individual consumer's preference generating XML format, and document is stored into individual consumer's preference storage and management module.
Fig. 2 G is the illustrative view of functional configuration of user preference self-adaptation subsystem in Fig. 2 of the present invention, as shown in Figure 2 G, described user preference self-adaptation subsystem 25, catches the change of group/individual user preference by using user preference change/collision detection module 251 in time according to the change threshold of setting or change detection algorithm; Or when the user preference extracted by multi-source channel merges, detect that user preference is inconsistent in time and produce conflict or extract inconsistent between user preference and the preference of the explicit setting of user and conflicting of causing; And then utilize described user preference correcting module 252 to revise, and correction result is kept in user preference database.
Fig. 3 is the physical arrangement schematic diagram that in the embodiment of the present invention, mobile network user preference obtains system, there is shown and comprise operator, service content provider, the relation of terminal user under interior mobile network environment in described prototype system, the data message that described data storage and management subsystem record Virtual network operator and service content provider produce for terminal user's mobile device, subsystem and user preference self-adaptation subsystem is extracted by using user preference, in conjunction with context environmental, the data message in described data storage and management subsystem is processed, to obtaining group's preference and the user preference information of described terminal user.
Fig. 3 A is the base conditioning schematic flow sheet that in mobile network of the present invention, user preference obtains system, is further detailed the acquisition methods of mobile network user preference in prototype system of the present invention below in conjunction with Fig. 3:
With reference to figure 3A, information interactive process between each subsystem of described prototype system, comprises the steps:
Step 301 ~ 302: generate subsystem 21 from user interface activated user historical behavior and context, generate user's historical behavior data and user's historical behavior context data;
Step 303: described user's historical behavior and context are generated subsystem user's historical behavior of generation and context thereof are stored in data storage and management subsystem 22;
Step 304 ~ 305: data mining subsystem 23 obtains user's historical behavior and context data thereof from data storage and management subsystem 22, carry out user clustering respectively and user behavior historical context calculates;
Step 306 ~ 313: user preference extracts subsystem 24 and extracts group user preference according to user clustering result, and then group user preference and user's historical behavior context result of calculation are carried out fusion calculation, to extract individual consumer's preference, and be stored into data storage and management subsystem 22;
Step 314: select to be applicable to the mobile network service type of this user at current context need for environment according to user preference, current context perception information, and recommend user;
Step 315: recommended mobile network service and content thereof are presented to user by user interface, and feed back;
Step 316 ~ 317: user preference self-adaptation subsystem 25 is according to user feedback, detect the change of user's historical behavior and context data thereof, or detect and extracted conflicting of user preference and real user demand, then adaptive correction is carried out to the user preference extracted, and through revised Output rusults, namely by user interface, recommended mobile network service and content thereof will be presented to user.
Fig. 4 is historical behavior context generative process schematic diagram in prototype system of the present invention, user's historical behavior context generation module 212 is using the Output rusults of user's historical behavior generation module 211 as input, first for user generates the personal information of user, then the rise time in units of mobile network service, reproducing device, and then generate position (scope of the customer location of the personal information decision of user), next will generating the activity (the generation result of User Activity depends on the positional information of user) of user, is finally that the volume of services of user distributes.Not separate above in each generative process, also Existence dependency and restriction relation between each agency, and finally generate user's historical behavior contextual information.As shown in Figure 4, this process specifically comprises:
Step 401: start historical behavior context and generate parent reason and each sub agent, receive the control message of superior agency (user's historical behavior generates agency).
Step 402: historical behavior context generates parent reason and obtain user's historical behavior data from database.
Step 403: user profile generates agency and receives the control message that historical behavior context generates parent reason, the variable of control message transmission is that user ID kimonos is make sure and used total amount.
Step 404: user profile generates the user profile that agency generates user, the user profile of generation is stored in user message table, and transmit control message to subscriber equipment generation agency.
Step 405: use subscriber equipment to generate agency and generate subscriber equipment for each user, and transmit control message to time generation agency.
Step 406: service time generates agency, is the time context of every class service creation user, makes the time that each record correspondence one is different, and transmits control message to position generation agency.
Step 407: generation agency in place to use is the place context of every class service creation user, makes each use the corresponding user position of record, and transmits control message to User Activity generation agency.
Step 408: use User Activity generates agency, is the action message context of every class service creation user historical behavior, makes each use all corresponding a kind of User Status of record, and transmits control message to volume of services generation agency.
Step 409: use volume of services assignment agent, user's historical behavior volume of services of each user is distributed, in the per a period of time allowing the volume of services of various mobile network service be assigned in 30 days of one month, and contextual for band user's historical behavior is stored in user's historical behavior context database.It should be noted that, every sub-distribution volume of services of a day/24 hours, the subscriber equipment that then transmits control message generates agency, the circulation that the volume of services next time performing this user distributes.
Step 410: volume of services assignment agent transmit control message subscriber equipment generate agency.
Step 411: the time generates after agency detects that the user behavior of this user one month/30 day is assigned, transmitting control message generates parent reason to historical behavior context, and the user's historical behavior context data performing next user generates.
In addition, in user's historical behavior context data generative process, use multi-thread concurrent control technology, each circulation generates the data of N (getting 5 ~ 10) individual user.
Fig. 5 is the schematic flow sheet based on individual consumer's interest-degree computing method of context calculating in prototype system of the present invention, using user's historical behavior context data as input, by carrying out reasoning to Context Knowledge and utilizing bayesian theory to calculate user's historical behavior context, solve various context probability when user's historical behavior occurs, final output packet is containing individual consumer's interest-degree of contextual information, as shown in Figure 5, this process comprises:
Step 501: the case data information extracting user's historical behavior context computing module;
Step 502: read user's historical behavior context database, read processed contextual information;
Here, described contextual information is the contextual information of raw information after the process of domain body obtained from sensor, and now contextual information is no longer original numeral, but a kind of semantic information;
Step 503: mate according to read user's historical behavior contextual information inquiry case database, check whether case exists, if case exists, then performs step 504; Otherwise, perform step 505;
Step 504: modify to the corresponding case in case library, then performs step 506;
Step 505: create a new case, then performs step 506;
Step 506: by the amendment made case or the new case created stored in case database, then perform step 507;
Step 507: judge whether that contextual information has read complete, if do not read, then return and perform step 501 continuation reading contextual information, until described contextual information reads complete, case library has created complete; Otherwise, perform step 508;
Step 508: terminate case extraction flow process and start context calculation process, perform step 509;
Step 509: read case information in case library, by calculating the precondition value in Bayesian network, then performs step 510;
Step 510: the conditional value obtained according to previous step, calculates the value of the Bayesian network of each subnet of separation, namely draws the daily behavior custom of user, then performs step 511;
Step 511: by the value of each subnet that draws stored in database;
Step 512: the service probability of use value calculating user according to the value of each precondition and each subnet;
Step 513: use the service probability of use value of clustering algorithm to user to carry out cluster analysis, so that probable value is divided into different grades, to extract individual consumer's interest-degree;
Step 514: by extracted individual consumer's interest-degree stored in individual consumer's interest-degree database;
Step 515: terminate computation process.
Fig. 6 is the user preference extracting method schematic flow sheet that fusion collaborative filtering in prototype system of the present invention and context calculate, model is obtained known with reference to the user preference shown in Fig. 1, the method is for input with user's historical behavior and user's historical behavior contextual information, utilize the collaborative filtering method based on clustering algorithm to calculate user's historical behavior, and extract group user preference; Utilize historical context computing method to calculate user's historical behavior context, and extract individual consumer's interest-degree; Then by based on collaborative filtering method group user preference and based on context calculate individual consumer's interest-degree carry out fusion calculation, thus predict and extract different user to the preference of different mobile network service under different context environmental, i.e. individual consumer's preference information.With reference to figure 6, the method comprises the steps:
Step 601: obtain individual consumer's interest-degree and group of subscribers preference, travel through each user under often kind of context environmental to the interest-degree that every class is served;
Step 602: judge whether user has traveled through, if do not traveled through, has then performed step 603; Otherwise, if execute, then perform step 613;
Step 603: judge whether the traversal of serving in certain class for certain individual consumer completes, if complete, returns step 602; Otherwise, perform step 604;
Step 604; Judge whether complete for the context traversal of described individual consumer under the service of certain class, if complete, return step 603; Then, step 605 is performed;
Step 605: whether the interest-degree judging described individual consumer is zero, if so, then performs step 606; Otherwise, perform step 611;
Step 606: the every other user traveling through this individual consumer place group to the interest-degree of such service, and obtains this user place group identification under this kind of context environmental, then performs step 607;
Step 607: calculate described every other user mean value to the interest-degree of such service under this kind of context environmental; Mean value described here is the mean value of the interest-degree of nonzero value, i.e. effective mean value;
Step 608: judge whether described effective mean value is zero, if so, then performs step 609; Then, step 610 is performed;
Step 609: individual consumer's preference value is the preference value that this user place group serves such, then performs step 612;
Step 610: individual consumer's preference value is effective mean value; Then step 612 is performed;
Step 611: individual consumer's preference value is the interest-degree of this individual consumer, then performs step 612;
Step 612: individual consumer's preference be stored in corresponding database, then returns and performs step 604;
Step 613: terminate this ergodic process.
The above, be only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.
Claims (6)
1. the acquisition system of user preference in a mobile network, it is characterized in that, described acquisition system comprises user's historical behavior and context generates subsystem (21), data storage and management subsystem (22), data mining subsystem (23), user preference extraction subsystem (24) and user preference self-adaptation subsystem (25); Wherein:
User's historical behavior and context generate subsystem (21), in order to complete mobile subscriber's historical behavior and the contextual data genaration of user's historical behavior; Described user's historical behavior and context generate subsystem (21) and comprise further: user's historical behavior generation module (211) and user's historical behavior context generation module (212); Wherein, described user's historical behavior generation module (211), for realizing the data genaration function of user's historical behavior, its Output rusults is the data source of group user preference extraction; Described user's historical behavior context generation module (212), for realizing the contextual data genaration function of user's historical behavior, its Output rusults is the data source that user's historical behavior context calculates;
Data storage and management subsystem (22), in order to the storage of completing user historical behavior data, user's historical behavior contextual information, user preference information, and for the tables of data relevant by user message table, user's historical behavior table, user's historical behavior contextual information table, group user preference table, individual consumer's preference table and user interface, Various types of data information is managed; Described contextual information, comprises customer location, time, environmental parameter, contiguous equipment and personnel, user activity information;
Data mining subsystem (23), calculates in order to completing user cluster and user's historical behavior context; Described data mining subsystem (23), comprises user clustering module (231) and user's historical behavior context computing module (232) further; Wherein, described user clustering module (231), based on the use amount of user to Information Mobile Service, by using K-Means clustering algorithm, all users be divided in the individual different cluster of K, make the user's similarity in same cluster higher, the user's similarity in different cluster is lower, after cluster analysis terminates, each user has a cluster labelled notation; Described user's historical behavior context computing module (232), for calculating user's historical behavior context, in the hope of going out individual consumer to contextual interest-degree in a certain respect;
User preference extracts subsystem (24), for the result of calculation according to described data mining subsystem (23), extracts group user preference and individual consumer's preference information, and outputs to described data storage and management subsystem (22); And
User preference self-adaptation subsystem (25), in order to the change/collision detection of completing user preference, user preference correction, and is kept at the result of testing result or correction in data storage and management subsystem (22).
2. the acquisition system of user preference in mobile network according to claim 1, is characterized in that, described user preference extracts subsystem (24), comprises group user preference extraction module (241) and individual consumer's preference extraction module (242) further; Wherein,
Described group user preference extraction module (241), for calculating group user cluster result and group user historical behavior, exports group user preference information;
Described individual consumer's preference extraction module (242), for carrying out fusion calculation to group user preference and user's historical behavior context result of calculation, exports individual consumer's preference information.
3. the acquisition methods of user preference in mobile network, it is characterized in that, the method comprises:
A, utilize user's historical behavior and context to generate subsystem (21) to generate user's historical behavior data and user's historical behavior context data, and described data are kept in data storage and management subsystem (22);
B, from described data storage and management subsystem (22), obtain user's historical behavior and context data by data mining subsystem (23), and carry out user clustering respectively and user behavior historical context calculates; Described data mining subsystem (23), comprises user clustering module (231) and user's historical behavior context computing module (232); Be specially: utilize user clustering module (231), based on the use amount of user to Information Mobile Service, by using K-Means clustering algorithm, all users are divided in the individual different cluster of K, make the user's similarity in same cluster higher, user's similarity in different cluster is lower, and after cluster analysis terminates, each user has a cluster labelled notation; And utilize user's historical behavior context computing module (232) to calculate user's historical behavior context, obtain individual consumer to contextual interest-degree in a certain respect;
C, extract subsystem (24) by user preference again and extract group user preference according to user clustering result, and then group user preference and user's historical behavior context result of calculation are carried out fusion calculation, to extract individual consumer's preference, and user preference information is stored in described data storage and management subsystem (22).
4. the acquisition methods of user preference in mobile network according to claim 3, is characterized in that, comprise further after described step C:
User preference self-adaptation subsystem (25) comprises user preference change/collision detection module (251) and user preference correcting module (252); Described user preference self-adaptation subsystem (25) is according to user feedback, detect the change of user's historical behavior and context data thereof, or detect and extracted conflicting of user preference and real user demand, then adaptive correction is carried out to the user preference extracted, and user will be presented to through revised Output rusults by mobile network service and content thereof.
5., based on individual consumer's interest-degree computing method that context calculates in the acquisition system of user preference described in claim 1, it is characterized in that, the method comprises:
The case data information of A, extraction user historical behavior context computing module, reads user's historical behavior context database, reads processed contextual information;
B, according to read user's historical behavior contextual information inquiry case database mate, check whether case exists, if case exist, then perform step C; Otherwise, perform step D;
C, the corresponding case in case database to be modified, then perform step e;
D, by the amendment made case or the new case created stored in case database, then perform step e;
E, to judge whether that contextual information has read complete, if do not read, then return and perform steps A and continue to read contextual information, until described contextual information reads complete, now case database creates complete; Otherwise, perform step F;
F, end case extract flow process, start context calculation process, read case information in case database, by calculating the precondition value in Bayesian network, and according to the conditional value obtained, calculate the value of the Bayesian network of each subnet of separation, draw the daily behavior custom of user, then by the value of each subnet that draws stored in higher-layer contexts database;
G, calculate the service probability of use value of user according to the value of each conditional value of having obtained and each subnet, the service probability of use value of clustering algorithm to user is used to carry out cluster analysis, so that probable value is divided into different grades, to extract individual consumer's interest-degree, and by extracted individual consumer's interest-degree stored in individual consumer's interest-degree database.
6. merge the user preference extracting method that collaborative filtering and context calculate in the acquisition system of user preference described in claim 1, it is characterized in that, the method comprises:
A, acquisition individual consumer's interest-degree and group of subscribers preference, travel through each user under often kind of context environmental to the interest-degree that every class is served;
B, judge whether user has traveled through, if do not traveled through, then perform step c; Otherwise, terminate this ergodic process;
C, judge whether the traversal of serving in certain class for certain individual consumer completes, if complete, returns step b; Otherwise, perform steps d;
D, judge for described individual consumer certain class service under context traversal whether complete, if complete, return step c; Otherwise, perform step e;
Whether e, the interest-degree judging described individual consumer are zero, if so, then perform step f; Otherwise, perform step k;
F, travel through this individual consumer place group every other user under this kind of context environmental to the interest-degree of such service, and obtain this user place group identification, then perform step g;
G, calculate described every other user effective mean value of interest-degree to such service under this kind of context environmental;
H, judge whether described effective mean value is zero, if so, then performs step I; Otherwise, perform step j;
I, individual consumer's preference value are the preference value that this user place group serves such, then perform step l;
J, individual consumer's preference value are effective mean value; Then step l is performed;
K, individual consumer's preference value are the interest-degree of this individual consumer, then perform step l;
L, individual consumer's preference is stored in corresponding database, then returns execution steps d.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910238504.6A CN102075851B (en) | 2009-11-20 | 2009-11-20 | Method and system for acquiring user preference in mobile network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910238504.6A CN102075851B (en) | 2009-11-20 | 2009-11-20 | Method and system for acquiring user preference in mobile network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102075851A CN102075851A (en) | 2011-05-25 |
| CN102075851B true CN102075851B (en) | 2015-01-07 |
Family
ID=44034166
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910238504.6A Expired - Fee Related CN102075851B (en) | 2009-11-20 | 2009-11-20 | Method and system for acquiring user preference in mobile network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102075851B (en) |
Families Citing this family (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102890689B (en) * | 2011-07-22 | 2017-06-06 | 北京百度网讯科技有限公司 | The method for building up and system of a kind of user interest model |
| CN103093370A (en) * | 2011-11-03 | 2013-05-08 | 阿里巴巴集团控股有限公司 | Method and device for commodity information launch |
| CN102421062B (en) * | 2011-12-01 | 2014-05-07 | 中国联合网络通信集团有限公司 | Method and system for pushing application information |
| CN103164474B (en) * | 2011-12-15 | 2016-03-30 | 中国移动通信集团贵州有限公司 | A kind of method that data service is analyzed |
| CN103377189A (en) * | 2012-04-12 | 2013-10-30 | 南京财经大学 | Method for collecting ground reference information of grain on basis of intelligent mobile terminals |
| CN103488525A (en) * | 2012-06-08 | 2014-01-01 | 诺基亚公司 | Determination of user preference relevant to scene |
| US9003025B2 (en) * | 2012-07-05 | 2015-04-07 | International Business Machines Corporation | User identification using multifaceted footprints |
| CN103678417B (en) * | 2012-09-25 | 2017-11-24 | 华为技术有限公司 | Human-machine interaction data treating method and apparatus |
| CN102855333A (en) * | 2012-09-27 | 2013-01-02 | 南京大学 | Service selection system based on group recommendation and selection method thereof |
| CN102915484A (en) * | 2012-10-12 | 2013-02-06 | 重庆亚德科技股份有限公司 | Intelligent predetermined plan system based on collaborative filtering |
| KR102183550B1 (en) * | 2013-03-13 | 2020-11-27 | 워런 존 패리 | A method of, and a system for, analysing data relating to an individual |
| CN103218400B (en) * | 2013-03-15 | 2017-04-05 | 北京工业大学 | Based on link and network community user group's division methods of content of text |
| CN103745384B (en) * | 2013-12-31 | 2017-06-06 | 北京百度网讯科技有限公司 | A kind of method and device for providing information to user equipment |
| CN105450598A (en) * | 2014-08-14 | 2016-03-30 | 上海坤士合生信息科技有限公司 | Information identification method, information identification equipment and user terminal |
| US10354206B2 (en) * | 2014-10-02 | 2019-07-16 | Airbnb, Inc. | Determining host preferences for accommodation listings |
| CN105718471A (en) * | 2014-12-03 | 2016-06-29 | 中国科学院声学研究所 | User preference modeling method, system, and user preference evaluation method and system |
| CN105095909A (en) * | 2015-07-13 | 2015-11-25 | 中国联合网络通信集团有限公司 | User similarity evaluation method and apparatus for mobile network |
| CN105354339B (en) * | 2015-12-15 | 2018-08-17 | 成都陌云科技有限公司 | Content personalization providing method based on context |
| CN105956009B (en) * | 2016-04-21 | 2019-09-06 | 深圳大数点科技有限公司 | A method of do something for the occasion in real time content matching and push |
| CN108874812B (en) | 2017-05-10 | 2021-12-10 | 腾讯科技（北京）有限公司 | Data processing method, server and computer storage medium |
| CN107357833B (en) | 2017-06-21 | 2020-05-26 | Oppo广东移动通信有限公司 | Data processing method and related product |
| CN107590224B (en) * | 2017-09-04 | 2021-11-30 | 北京京东尚科信息技术有限公司 | Big data based user preference analysis method and device |
| CN108230094B (en) * | 2017-12-22 | 2021-11-16 | 金瓜子科技发展（北京）有限公司 | Vehicle recommendation method and device |
| CN108874959B (en) * | 2018-06-06 | 2022-03-29 | 电子科技大学 | User dynamic interest model building method based on big data technology |
| CN109788056A (en) * | 2019-01-10 | 2019-05-21 | 四川新网银行股份有限公司 | User's theme message method for pushing and system based on clustering |
| CN111787570B (en) * | 2020-06-19 | 2023-11-03 | 深圳市有方科技股份有限公司 | Data transmission method and device of Internet of things equipment and computer equipment |
| CN112770181A (en) * | 2021-01-12 | 2021-05-07 | 贵州省广播电视信息网络股份有限公司 | Quick verification system and method for recommended content of family group |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1573725A (en) * | 2003-06-20 | 2005-02-02 | 英特尔公司 | Method, apparatus and system for enabling context aware notification in mobile devices |
| CN101071424A (en) * | 2006-06-23 | 2007-11-14 | 腾讯科技（深圳）有限公司 | Personalized information push system and method |
| CN101079824A (en) * | 2006-06-15 | 2007-11-28 | 腾讯科技（深圳）有限公司 | A generation system and method for user interest preference vector |
-
2009
- 2009-11-20 CN CN200910238504.6A patent/CN102075851B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1573725A (en) * | 2003-06-20 | 2005-02-02 | 英特尔公司 | Method, apparatus and system for enabling context aware notification in mobile devices |
| CN101079824A (en) * | 2006-06-15 | 2007-11-28 | 腾讯科技（深圳）有限公司 | A generation system and method for user interest preference vector |
| CN101071424A (en) * | 2006-06-23 | 2007-11-14 | 腾讯科技（深圳）有限公司 | Personalized information push system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102075851A (en) | 2011-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102075851B (en) | Method and system for acquiring user preference in mobile network | |
| Amato et al. | SOS: a multimedia recommender system for online social networks | |
| Dhelim et al. | Personality-aware product recommendation system based on user interests mining and metapath discovery | |
| Costa-Montenegro et al. | Which App? A recommender system of applications in markets: Implementation of the service for monitoring users’ interaction | |
| Roh et al. | The collaborative filtering recommendation based on SOM cluster-indexing CBR | |
| CN106503015A (en) | A kind of method for building user's portrait | |
| CN103106285B (en) | Recommendation algorithm based on information security professional social network platform | |
| CN106504099A (en) | A kind of system for building user's portrait | |
| EP3070661A1 (en) | System and method for providing context driven hyper-personalized recommendation | |
| CN103970891B (en) | A kind of user interest information querying method based on situation | |
| CN103678647A (en) | Method and system for recommending information | |
| CN103154928A (en) | Network server arrangement for processing non-parametric, multi-dimensional, spatial and temporal human behavior or technical observations measured pervasively, and related method for the same | |
| CN101770520A (en) | User interest modeling method based on user browsing behavior | |
| CN103295145A (en) | Mobile phone advertising method based on user consumption feature vector | |
| CN102999586A (en) | Method and device for recommending website | |
| CN103544623A (en) | Web service recommendation method based on user preference feature modeling | |
| CN110704574B (en) | Banking business demand asset management method and system | |
| Lin et al. | Premise: Personalized news recommendation via implicit social experts | |
| CN102402594A (en) | Rich media individualized recommending method | |
| CN105787066A (en) | Digital content distribution system based on total analysis | |
| KR101475439B1 (en) | System and method for providing of interesting information optimized for user | |
| Li et al. | A Personalization Recommendation Algorithm for E-Commerce. | |
| Li et al. | A two-tier service filtering model for web service QoS prediction | |
| Qiu et al. | Design of multi-mode e-commerce recommendation system | |
| Kwak | A study on the evolution of post-smartphone technologies in the 5G technology environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150107 Termination date: 20151120 |
|
| EXPY | Termination of patent right or utility model |