CN107194515A - Determine user's current behavior and the method and apparatus for predicting user view - Google Patents
Determine user's current behavior and the method and apparatus for predicting user view Download PDFInfo
- Publication number
- CN107194515A CN107194515A CN201710396359.9A CN201710396359A CN107194515A CN 107194515 A CN107194515 A CN 107194515A CN 201710396359 A CN201710396359 A CN 201710396359A CN 107194515 A CN107194515 A CN 107194515A
- Authority
- CN
- China
- Prior art keywords
- user
- information
- current
- analysis model
- network
- 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.)
- Pending
Links
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
Abstract
The present invention provides the method and apparatus for determining user's current behavior and predicting user view, and the method for prediction user view therein includes:Obtain the current characteristic information of user;User identity information in the current characteristic information of user determines the corresponding Analysis model of network behaviors of user;By in the corresponding Analysis model of network behaviors of information input user of sign user's current state in the current characteristic information of user;According to the current behavior acquisition of information of the user of the corresponding Analysis model of network behaviors output of the user network data related to the current behavior information of user;According to the current behavior for determining user to the result of the current behavior information of user using network data;Wherein, the current characteristic information of user is stored as the history feature information of the user of the model parameter for setting and updating the corresponding Analysis model of network behaviors of user.The present invention can make, using the actual demand for preferably meeting user, to improve life cycle and the competitiveness of application, reduce the cost of application.
Description
Technical field
The present invention relates to Internet technology, more particularly to a kind of method for determining user's current behavior, determine that user is current
The device of behavior, the method for predicting user view and the device for predicting user view.
Background technology
With the fast development and popularization of Internet technology especially development of Mobile Internet technology, the intelligence of network is had access to
Mobile device (such as intelligent mobile phone and tablet personal computer) has become many people article indispensable with oneself, and people utilize it
Intelligent mobile equipment can realize mail transmission/reception, instant message interaction and network access etc. anywhere or anytime.
Inventor is had found in process of the present invention is realized, the application in Intelligent mobile equipment is mountable at present
The species and quantity of (Application, APP) are all increasingly enriched, and how to make to apply the concern for winning user with liking, to carry
The life cycle of height application and viscosity, reduce the cost of application, all right and wrong for application and development side and application operation side
It is often important.
The content of the invention
It is an object of the invention to provide a kind of method and apparatus for determining user's current behavior and predicting user view.
According to the first aspect of the invention there is provided a kind of method of pre-determining user current behavior, wherein, this method
Mainly include the following steps that:Obtain the current characteristic information of user;User's mark in the current characteristic information of the user
Know information and determine the corresponding Analysis model of network behaviors of the user;Sign user in the current characteristic information of the user is current
The information of state is inputted in the corresponding Analysis model of network behaviors of the user;According to the corresponding Analysis model of network behaviors output of the user
User the current behavior acquisition of information network data related to the current behavior information of the user;According to the utilization net
Network data determine the current behavior of the user to the result of the current behavior information of the user;Wherein, the user
Current characteristic information be stored as the history feature information of the user, and the history feature information is used to set and update
The model parameter of the corresponding Analysis model of network behaviors of the user.
According to the second aspect of the invention there is provided it is a kind of predict user view method, wherein, this method include with
Lower step:Obtain the current characteristic information of user;User identity information in the current characteristic information of the user is determined
The corresponding Analysis model of network behaviors of the user;By the information of sign user's current state in the current characteristic information of the user
Input in the corresponding Analysis model of network behaviors of the user;The user's exported according to the corresponding Analysis model of network behaviors of the user works as
Move ahead as the acquisition of information network data related to the current behavior information of the user;According to the utilization network data to institute
The result for stating the current behavior information of user determines the current behavior of the user;It is pre- according to the current behavior of the user
Survey the current intention of the user;Wherein, the current characteristic information of the user is stored as the history feature letter of the user
Breath, and model parameter of the history feature information for setting and updating the corresponding Analysis model of network behaviors of the user.
According to the third aspect of the present invention there is provided a kind of device for determining user's current behavior, wherein, the device bag
Include:Obtain characteristic information module, the current characteristic information for obtaining user;Analysis model module is determined, for according to described
User identity information in the current characteristic information of user determines the corresponding Analysis model of network behaviors of the user;Information processing mould
Block, for the information of sign user's current state in the current characteristic information of the user to be inputted into the corresponding row of the user
For in analysis model;Network data model is obtained, for the user's according to the corresponding Analysis model of network behaviors output of the user
The current behavior acquisition of information network data related to the current behavior information of the user;Current behavior module is determined, is used for
According to the current behavior for determining the user to the result of the current behavior information of the user using the network data;
Wherein, the current characteristic information of the user is stored as the history feature information of the user, and the history feature information
Model parameter for setting and updating the corresponding Analysis model of network behaviors of the user.
According to the fourth aspect of the present invention there is provided a kind of device for predicting user view, wherein, the device includes:
Obtain characteristic information module, the current characteristic information for obtaining user;Analysis model module is determined, for according to the user
Current characteristic information in user identity information determine the corresponding Analysis model of network behaviors of the user;Message processing module, is used
Divide in the information of sign user's current state in the current characteristic information of the user is inputted into the corresponding behavior of the user
Analyse in model;Network data model is obtained, for the current of the user according to the corresponding Analysis model of network behaviors output of the user
Behavioural information obtains the network data related to the current behavior information of the user;Current behavior module is determined, for basis
The current behavior of the user is determined to the result of the current behavior information of the user using the network data;It is intended to
Prediction module, the current intention for predicting the user according to the current behavior of the user;Wherein, the user's is current
Characteristic information is stored as the history feature information of the user, and the history feature information is used to set and update the use
The model parameter of the corresponding Analysis model of network behaviors in family.
Compared with prior art, the present invention has advantages below:Current characteristic information of the invention by obtaining user, one
Aspect can accumulate the history feature information for obtaining the substantial amounts of user, so as to be built using the history feature information of the user
Found the Analysis model of network behaviors of user institute exclusive (not for other users all), it is possible to the history feature information of the utilization user
Constantly adjustment (updates) model parameter of the Analysis model of network behaviors corresponding to the user, makes the behavior point corresponding to the user
Analysis model can press close to the real life behavior of the user as far as possible all the time, on the other hand can utilize the corresponding row of the user
The information of sign user's current state in the current characteristic information of the user is calculated for analysis model, makes the present invention can
The current behavior information of user is timely obtained with the information based on the Analysis model of network behaviors output corresponding to the user;Pass through profit
The current behavior information of user with related network data to currently obtaining is verified, can more accurately judge to use
The current behavior at family;Due to knowing that user's current behavior can make using corresponding service is targetedly provided a user, such as
The current intention of user is predicted by the current behavior according to user, and then when application pushes new information to user, can
Meet its information being currently intended to be pushed to user, used this way it is possible to avoid the information pushed is given due to reasons such as redundancies
The problems such as puzzlement is brought at family, so as to increase viscosity of the user to application;It follows that the technical scheme energy that the present invention is provided
Enough make, using the actual demand of user is preferably met, to improve life cycle and the competitiveness of application, and then reduce should
Cost.
Brief description of the drawings
By reading the detailed description made to non-limiting example made with reference to the following drawings, of the invention is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is the method flow diagram of determination user's current behavior of the embodiment of the present invention one;
Fig. 2 is the method flow diagram of the prediction user view of the embodiment of the present invention two;
Fig. 3 is the schematic device of determination user's current behavior of the embodiment of the present invention three;
Fig. 4 is the schematic device of the prediction user view of the embodiment of the present invention four.
Same or analogous reference represents same or analogous part in accompanying drawing.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, the implementation below in conjunction with accompanying drawing to the present invention
Example is described in detail.
It should be mentioned that some exemplary embodiments are described as before exemplary embodiment is discussed in greater detail
The processing described as flow chart or method.Although operations are described as the processing of order by the flow chart of the present invention,
Many of which operation can be implemented concurrently, concomitantly or simultaneously.In addition, the order of operations can be pacified again
Row.The processing can be terminated when its operations are completed, it is also possible to the additional step being not included in accompanying drawing.Institute
State processing and can correspond to method, function, code, subroutine, subprogram etc..
The intelligent electronic device includes user equipment and the network equipment.Wherein, the user equipment includes but is not limited to
Computer, intelligent mobile phone and PDA etc.;The network equipment includes but is not limited to single network server, multiple networks clothes
Be engaged in device composition server group or based on cloud computing (a oud Computing) by a large amount of computers or the webserver
The cloud of composition, wherein, cloud computing is one kind of Distributed Calculation, and one be made up of the computer collection of a group loose couplings is super
Virtual machine.Wherein, the intelligent electronic device can be carried out with access network and with other intelligent electronic devices in network
Information exchange is operated.Wherein, the network that the intelligent electronic device is had access to include but is not limited to internet, wide area network,
Metropolitan Area Network (MAN), LAN, VPN etc..
It should be noted that the user equipment, the network equipment and network etc. are only for example, other are existing or from now on
The intelligent electronic device or network being likely to occur such as are applicable to the application, should also be included within the application protection domain,
And be incorporated herein by reference.
Describe below discussed method (some of them are illustrated by flow) embodiment can by hardware, software,
Firmware, middleware, microcode, form of hardware description language or its any combination are implemented.When with software, firmware, centre
When part or microcode are to implement, program code or code segment to implement necessary task can be stored in machine or meter
In calculation machine computer-readable recording medium (such as storage medium).(one or more) processor can implement necessary task.
Concrete structure and function detail disclosed herein are only representational, and are for describing showing for the application
The purpose of example property embodiment, still, the application can be implemented by many alternative forms, and be not interpreted as
It is limited only by the embodiments set forth herein.
Although it should be appreciated that may have been used term " first ", " second " etc. herein to describe unit,
But these units should not be limited by these terms.It is used for the purpose of using these terms by a unit and another unit
Make a distinction.For example, in the case of the scope without departing substantially from exemplary embodiment, it is single that first module can be referred to as second
Member, and similarly second unit can be referred to as first module.Term "and/or" used herein above include one of them or
Any and all combination of more listed associated items.
It should be appreciated that when a unit is referred to as " connecting " or during " coupled " to another unit, it can be straight
Connect and be connected or coupled to another described unit, there can also be temporary location.On the other hand, when a unit is referred to as
When " being directly connected to " or " direct-coupling " arrives another unit, then in the absence of temporary location.It should come in a comparable manner
Explain other words that be used to describing the relation between unit (for example, " between being in ... " is compared to " being directly in ... it
Between ", " with ... it is neighbouring " compared to " with ... it is directly adjacent to " etc.).
Term used herein above is just for the sake of description specific embodiment, without being intended to limit exemplary embodiment.
Unless clearly referred else in context, otherwise, singulative " one " used herein above, " one " also attempt to include again
Number.It is to be further understood that term " comprising " used herein above and/or "comprising" define stated feature, integer, step
Suddenly, the presence of operation, unit and/or component, and do not preclude the presence or addition of other one or more features, integer, step
Suddenly, operation, unit, component and/or its combination.
It should further be mentioned that in some replaces realization modes, the function/action being previously mentioned can be according to different from attached
The order indicated in figure occurs.For example, depending on involved function/action, the two width figures shown in succession actually may be used
Substantially simultaneously to perform or can perform in a reverse order sometimes.
Technical scheme is described in further detail below in conjunction with the accompanying drawings.
Embodiment one, the method for determining user's current behavior.
Fig. 1 is the method flow diagram of determination user's current behavior of the present embodiment.In the flow chart shown in Fig. 1, this reality
Applying the method for determination user's current behavior of example includes:Step S100, step S110, step S120, step S130 and step
S140。
Method described in the present embodiment is performed in intelligent electronic device, and this method is typically in network side
Being performed in intelligent electronic device (the server network equipment as being arranged at high in the clouds), certainly, the present embodiment is also not excluded for this
The possibility that method is performed in the intelligent electronic device (such as intelligent mobile phone or tablet personal computer user equipment) of user side
Property (model therein can by network side the network equipment train, and be issued to user side).It is true that the present embodiment does not limit realization
Determine the specific manifestation form of the intelligent electronic device of the method for user's current behavior, i.e. the present embodiment do not limit determination user it is current
The hardware environment that the method for behavior is applicable.
Each step in Fig. 1 is described in detail respectively below.
S100, the current characteristic information for obtaining user.
Specifically, the current characteristic information of the user in the present embodiment is the user in current time or current slot
Characteristic information;The current characteristic information of user in the present embodiment can be stored as the history feature information of user;
That is, the current characteristic information of user and the history feature information of user are the characteristic information of user, in working as user
Preceding characteristic information lose its it is ageing when, the current characteristic information of user turns into the history feature information of user.
The characteristic information of user in the present embodiment refers to the information of the characteristic feature for describing user, is such as used to describe
The static nature of user and the information of behavioral characteristics.
The static nature of user in the present embodiment is properly termed as user property in user property, and the present embodiment can be with
Including:The sex of user, the age bracket (or year of birth generation of user) of user, the occupational group of user, the address of user, use
The business address at family, the incommutation instrument of user, the work and rest rule of user, the income level of user, the education water of user
Behavior preference of flat and user etc..
The behavioral characteristics of user in the present embodiment are properly termed as the state of user, and the state of the user in the present embodiment
It can specifically include:Environment that the motion state of user, user are in and the particular location where user etc..
The characteristic information of user in the present embodiment (i.e. believe by the current characteristic information of user and the history feature of user
Breath) generally include:The information of user's current state is characterized, and characterizes information sign user's current state of user's current state
Information can be specially sensor collection information, the satellite positioning information of user equipment in user equipment and based on user
Equipment microphone collection audio-frequency information in one or more.Sensor collection information in above-mentioned user equipment can be with
The information of gyroscope collection specially in user equipment, the information of the gravity sensor collection in user equipment and user set
One or more in the information of acceleration transducer collection in standby.The satellite positioning information of above-mentioned user equipment can have
Body is the location information based on GPS (G Iobal Position System, global positioning system) of user equipment, certainly, is used
The satellite positioning information of family equipment can also be specially location information based on big-dipper satellite alignment system of user equipment etc..On
The audio-frequency information for stating the microphone collection of user equipment be usually user sound in the environment.
The characteristic information of user in the present embodiment (i.e. believe by the current characteristic information of user and the history feature of user
Breath) generally also include:User's mark, the telephonic communication record in user equipment, the message registration of user, log information, Yong Hushe
For any one in each application message and user equipment model of middle installation or multiple.Above-mentioned user's mark can be specific
IMSI (the International Mobile Subscriber of Mobile Directory Number or user for user
Identification Number, international mobile subscriber identity) etc..The message registration of above-mentioned user is generally specially most
The message registration of user in nearly predetermined amount of time.Above-mentioned log information can be specially user using each application historical record,
When the used network address and user carry out network access using browser when user carries out network access using browser
Search keyword inputted etc..Each application message installed in above-mentioned user equipment can be specially to install in user equipment
The list information of application.
Under normal conditions, the content that the characteristic information of the user of the present embodiment is included can comprehensive as far as possible one
A bit, so as to being conducive to accurately determining user's current behavior.In addition, it is necessary to special instruction, the spy of above-mentioned user
Reference ceases the content that the current characteristic information that included content is both user is included, and is also the history feature information institute of user
Comprising content.
The present embodiment method by the case that user equipment is performed, user equipment can be with timing or not timing
Mode obtains the current characteristic information of user, and such as microphone of the triggering user equipment of timing or not timing carries out audio and adopted
Collection, with obtain the user in short time period sound in the environment;For another example in the collection user equipment of timing or not timing
The attitude information etc. that produces of gyroscope;For another example, the message registration for being obtained from preceding once acquisition user of timing or not timing
And the message registration and log information of the user produced by after the log information time.In addition, the present embodiment should obtained
In the case of the permission of user, the current characteristic information of user can be just got.
The present embodiment method by be arranged at network side the network equipment perform in the case of, user equipment can be used
The current characteristic information that the mode of timing or not timing is obtained user is reported to network side, so that being arranged at network side
The network equipment get the current characteristic information of user.The present embodiment not restricting user equipment to network side report of user work as
The specific implementation of preceding characteristic information.
S110, the corresponding behavioural analysis mould of user determined according to the user identity information in the current characteristic information of user
Type.
Specifically, the present embodiment would generally be set in Analysis model of network behaviors, i.e. the present embodiment respectively for multiple users in advance
It is usually provided with multiple Analysis model of network behaviors, and each Analysis model of network behaviors one user of correspondence, and different behavioural analyses
Model generally corresponds to different users, i.e., different users has different Analysis model of network behaviors.The present embodiment can will be each
Individual Analysis model of network behaviors it is corresponding with user identity information storage so that the present embodiment get user current characteristic information it
Afterwards, can be searched in the information prestored with the user identity information in the current characteristic information of user match it is corresponding
Relation, and it regard the Analysis model of network behaviors in the corresponding relation matched as the corresponding Analysis model of network behaviors of the user.
For a specific user, the present embodiment is to be set using the history feature information of the user for the user
Its corresponding Analysis model of network behaviors, a detailed process for setting Analysis model of network behaviors for the user can be:
First, multiple basic models, and a kind of attribute type of each basic model correspondence are previously provided with the present embodiment
User, and the user of different basic model correspondence different attribute type, such as the present embodiment can be for working from 9am to 5pm
This attribute type of working clan sets a basic model, and for this attribute type of working clan using flexible job system
One basic model etc. is set;The sorting particles degree of user property type corresponding to basic model in the example above is thicker,
And in actual applications, the sorting particles degree of the user property type corresponding to the basic model of the present embodiment can be thinner, and
The mode classification of the attribute type of user can be set according to the actual requirements.
Secondly, after the substantial amounts of history feature information of a user is obtained, the present embodiment can be according to the user
Substantial amounts of history feature information determine the attribute of the user and the type belonging to the attribute of the user, such as by the way that this is used
In the user property disaggregated model that pre-sets of substantial amounts of history feature information input at family, thus the present embodiment can according to
The information of family attributive classification model output determines the type belonging to the attribute of the user.User property classification in the present embodiment
Model be learnt by the history feature information to a large number of users formed by, as the present embodiment can be using having supervision
Mode of learning learns to the history feature information of a large number of users, so as to form user property disaggregated model;It is above-mentioned to a large amount of
The result that the history feature information of user is learnt and the mode that user classifies are closely related.
Include in the user property of the present embodiment:The sex of user, the age bracket (or year of birth generation of user) of user,
The occupational group of user, the address of user, the business address of user, the incommutation instrument of user, the work and rest rule of user,
In the case of the Behavior preference of the income level of user, the level of education of user and user, user property disaggregated model can be with
According to log information (such as network access address and the search keyword of a large amount of (such as one month or some months) of a user
Deng) etc. determine that user likes the content browsed, and the satellite positioning information of substantial amounts of user equipment according to the user etc. is really
Place that the fixed user likes in leisure time etc., so as to be tentatively inferred to the sex of user, the age bracket of user with
And the level of education of user etc.;User property disaggregated model can be according to a large amount of (such as one month or some months) of a user
User equipment satellite positioning information determine user periodically in the residence time of some position, may thereby determine that out
The work and rest rule in the address of user, the business address of user and user etc., and the sensor collection information in user equipment
The auxiliary information for the work and rest rule for determining the address of user, the business address of user and user can be turned into;User property
Disaggregated model can according to a large amount of (such as one month or some months) telephone relations record (whether call frequent) of user,
The satellite positioning information (working time whether be commonly located at a small regional extent in etc.) and user equipment of user equipment
In the rough occupational group for determining user of sensor collection information etc., user property disaggregated model can exist according to user
The satellite positioning information of the user equipment of moving process can determine the friendship on and off duty of user between its address and business address
Logical instrument, and sensor collection information in user equipment can also turn into the auxiliary for determining the incommutation instrument of user
Information;The incommutation instrument of user and the address of user etc. also contribute to the further income level for determining user with
And the level of education of user etc.;User property disaggregated model can be according to the satellite fix of the substantial amounts of user equipment of a user
Information, the audio-frequency information of microphone collection based on user equipment and the work and rest rule of the above-mentioned user determined can be determined
Go out Behavior preference of user etc., restaurant that such as restaurant often gone at noon of user, user often go on day off, user are frequent
Shopping place that the frequency and time and user that the cinema that goes, user see a film often are gone etc., in addition, in user equipment
Sensor collection information can also turn into the auxiliary information for determining the Behavior preference of user.It above are only and user property is classified
Model determines one of the process of user property for example, the present embodiment, which does not limit user property disaggregated model, determines use
The specific implementation of family attribute.
As an example, the user property disaggregated model of the present embodiment can determine to use based on the partial content in user property
Type belonging to the attribute of family, such as user property disaggregated model can according to the sex of user, the age bracket of user, user occupation
Classification, the incommutation instrument of user, the income level of user and the level of education of user are classified to user property,
The present embodiment can pre-set multiple types, and the specific value of the parameter included according to each type according to the actual requirements
To determine the type belonging to the user property of active user, the present embodiment does not limit user property disaggregated model and determines user property
The specific implementation of affiliated type.
Again, the present embodiment, can be from the multiple bases pre-set after the type belonging to the user property is determined
The basic model with the type matching belonging to the user property is chosen in plinth model;I.e. the present embodiment is previously provided with multiple bases
Plinth model, and each basic model is to that should have a kind of user property type, so as to determine the class belonging to user property
The basic model that should be used is determined during type.
Finally, the type matching that the present embodiment can be according to belonging to the history feature information of the user be user property
Basic model sets personalized model parameter, so as to form the corresponding Analysis model of network behaviors of the user;As the present embodiment can profit
The above-mentioned basic model matched is trained with a large amount of history feature information of the user, so that the mould of the basic model
Shape parameter is the personalized model parameter set for the user, that is, the basic model after training is the corresponding behavior point of user
Analyse model.Personalized model parameter in the present embodiment can be the address based on user, the business address of user and user
Behavior preference etc. and the model parameter that sets.
The Analysis model of network behaviors of the present embodiment can be used includes the calculation by core of HMM such as HMM and GMHMM based on+HMM
Method) Analysis model of network behaviors, and identification sorting algorithm in the Analysis model of network behaviors based on+HMM can use " forwards algorithms ".
The present embodiment is no longer described in detail herein to the specific training process of Analysis model of network behaviors.
Special instruction is additionally needed, in the case where being successfully provided with its Analysis model of network behaviors for user, this
The current characteristic information for the user that embodiment is subsequently got can be by the history feature information storage as user, and subsequently stores
The history feature information of user still can be used for being trained the Analysis model of network behaviors of the user, constantly to correct
The personalized model parameter of the Analysis model of network behaviors of (adjust or adjust) user, makes the individual character of the Analysis model of network behaviors of user
The actual conditions with user can preferably be shown by changing model parameter.
S120, the corresponding behavior of information input user by sign user's current state in the current characteristic information of user
In analysis model.
Specifically, the information of sign user's current state in the current characteristic information of the user of the present embodiment can be specific
For sensor collection information, the satellite positioning information of user equipment in user equipment and Mike's elegance based on user equipment
The audio-frequency information of collection.
The present embodiment can be first to sign user's current state in the current characteristic information of user information arrange,
Then, then by the information of sign user's current state after arrangement input in the corresponding Analysis model of network behaviors of user.
Arranged to the information of sign user's current state in the current characteristic information of user one of the present embodiment
Specific example is:The present embodiment is in situation about being stored in the current characteristic information got in current characteristic information set
Under, the current characteristic information of the untreated user can be first read from current characteristic information set, and recognize what is read out
The type of each current characteristic information, the type of current characteristic information can be address list type, posture type (in such as user equipment
Sensor collection information belong to posture type), audio types (such as based on user equipment microphone collection audio-frequency information belong to
In audio types) and align_type (satellite positioning information of such as user equipment belongs to align_type);Afterwards, the present embodiment
The information for characterizing user's current state, and the different types of table that the same time will be corresponded to can be chosen according to the type identified
The information of requisition family current state is combined, so as to form at least one status information group by dimension of the time.
Sign user's current state in one specific example, the current characteristic information for the user that the present embodiment is got
Information can be expressed as the form shown in following formula (1):
[Motion data from tn-1 to tn]+[Sound data from tn-1 to tn]+[Location
data fromtn-1 to tn];Formula (1)
In above-mentioned formula (1), [*] represents one group of continuous data, and tn-1 represents the time of previous processing data, tn
The time of this processing data is represented, and there are multiple moment between tn-1 and tn to that there should be the current spy of the user got
Reference ceases;Motion data represent the sensor collection information in user equipment, and Sound data are represented based on user equipment
The audio-frequency information of microphone collection, Location data represent the satellite positioning information of user equipment;
The present embodiment can be by the information of sign user's current state in the current characteristic information shown in above-mentioned formula (1)
Arrange as the form shown in following formula (2):
[(Motion Type at tn-1, Sound Type at tn-1, Location Type at tn-1) ...
(Motion Type at tn, Sound Type at tn, Location Type at tn)] formula (2)
In above-mentioned formula (2), [*] represents one group of continuous data, and (*) represents one in one group of continuous data
Status information group, tn-1 represents the time of previous processing data, and tn represents the time of this processing data, and between tn-1 and tn
There are multiple moment to that there should be the current characteristic information of the user got;Motion dat a represent the biography in user equipment
Sensor gathers information, and Sound dat a represent the audio-frequency information of the microphone collection based on user equipment, Location dat a
Represent the satellite positioning information of user equipment.
From sign user's current state in above-mentioned formula (2), the current characteristic information of the user of the present embodiment
Information is recombined as multiple status information groups according to time maintenance.
The present embodiment can sequentially input each status information group according to the sequencing (i.e. time sequencing) of time dimension
In the corresponding Analysis model of network behaviors of user.Continuous precedent, the present embodiment can be first by (Motion Type at tn-1, Sound
Type at tn-1, Location Type at tn-1) in the corresponding Analysis model of network behaviors of input user, then, by tn-1 and
The status information group of each intermediate time between tn is sequentially input in the corresponding Analysis model of network behaviors of user, until by (Motion
Type at tn, Sound Type at tn, Location Type at tn).
S130, the current behavior acquisition of information of the user exported according to the corresponding Analysis model of network behaviors of user and above-mentioned user
The related network data of current behavior information.
Specifically, the current behavior information of the user in the present embodiment is the information for describing the current behavior of user,
And the current behavior of user is referred to as the current life-form structure of user or the event of user etc..The current behavior of user can
Be specially over/under on the road of class (using/driving/public transport mode of walking), removing certain destination (such as film
Institute) road on (using/driving/public transport mode of walking), having lunch (at the restaurant/family/office), in work
(/ bending over one's desk working/on the road gone on business just in session), taking a nap or going shopping etc..The present embodiment is not limited
The specific manifestation form of user's current behavior processed.
The Analysis model of network behaviors of user in the present embodiment can according to input sign user's current state information (such as
Sensor collection information, the satellite positioning information of user equipment in user equipment and the microphone collection based on user equipment
Audio-frequency information) obtain the probability of each possible user behavior, the Analysis model of network behaviors of such as user may for each
User behavior be utilized respectively pre-defined algorithm to input sign user's current state information carry out probability calculation, it is defeated to obtain
The probability of each possible user behavior of the information correspondence of the sign user's current state entered;The Analysis model of network behaviors of user exists
Calculating during the probability of each possible user behavior would generally go through to characterizing the information of user's current state, user
The many-side such as historical event part and log information carries out comprehensive consideration.The present embodiment does not limit the Analysis model of network behaviors of user to input
Sign user's current state information calculated implement process.
The present embodiment can obtain the network number of correlation in order that the current behavior of the user finally determined is more accurate
According to the current behavior information of the user exported with the Analysis model of network behaviors to user is verified, such as obtains the row with user
The related network data of the current behavior information of the probability highest user exported by analysis model, so as to for user
The current behavior information of probability highest user that is exported of Analysis model of network behaviors verified, obtain respectively for another example and user
The probability highest that is exported of Analysis model of network behaviors and secondary high user the related network data of current behavior information so that
The probability highest and the current behavior information of secondary high user that can be exported for the Analysis model of network behaviors of user are carried out respectively
Checking.
The specific example that the present embodiment obtains related network data is:It is defeated for the Analysis model of network behaviors of user
The current behavior information of the probability highest user gone out carries out word segmentation processing, the key in current behavior information to extract user
Word, then, searches the network data related to keyword, such as the present embodiment can profit using the keyword extracted from network
The network data related to keyword is searched for modes such as web crawlers.The present embodiment does not limit the Analysis model of network behaviors from user
The specific implementation of the current behavior information extraction keyword of the user exported and lookup and keyword phase from network
The specific implementation of the network data of pass.
S140, user determined to the result of the current behavior information of user according to the network data using above-mentioned acquisition
Current behavior.
Specifically, the present embodiment can be according to Analysis model of network behaviors of the above-mentioned related network data got to user
The current behavior information (the current behavior information of such as probability highest user) of the user of the corresponding probability exported verified,
Such as verifying the current behavior information of probability highest user, whether the corresponding network data with getting is consistent, if be consistent,
Then determine to be verified, so that the probability highest user's that the present embodiment can be exported the Analysis model of network behaviors of user works as
Preceding behavioural information as description user current behavior information;If be not consistent, it is determined that authentication failed, so that the present embodiment
The current behavior information for the probability highest user that the Analysis model of network behaviors of user can not be exported is worked as description user's
Move ahead the information for being.Above-mentioned to be consistent can be specially that the time parameter such as place is consistent, a specific example, in the row of user
The current behavior information of the probability highest user exported by analysis model represents that user is being gone on the road in the first gymnasium
In the case of, will be in A B months C day D if including certain singer by the network data accessed by keyword of the first gymnasium
When hold concert in the first gymnasium, then the present embodiment is determining that the current date is A B months C day, and current time faces
In the case of nearly D, it may be determined that this is verified, so that the present embodiment can accurately determine the current behavior of user
To go on the road in the first gymnasium, conversely, then the present embodiment can not accurately determine user current behavior whether be
Going on the road in the first gymnasium.
The current behavior for the user that the present embodiment is determined can be not only used for predicting user view, and can be used for
Other scenes, such as by recording life course of the user in the range of a period of time, can be presented to user when user checks
Or regularly it is presented to user etc..The present embodiment does not limit the concrete application scene of the current behavior of the user determined.
Embodiment two, the method for predicting user view.
Fig. 2 is the method flow diagram of the prediction user view of the present embodiment.
In the flow shown in Fig. 2, the method for the prediction user view of the present embodiment mainly includes:Step S200, step
S210, step S220, step S230, step S240 and step S250.
Method described in the present embodiment is performed in intelligent electronic device, and this method is typically in network side
Being performed in intelligent electronic device (the server network equipment as being arranged at high in the clouds), certainly, the present embodiment is also not excluded for this
The possibility that method is performed in the intelligent electronic device (such as intelligent mobile phone or tablet personal computer user equipment) of user side
Property (model therein can by network side the network equipment train, and be issued to user side).It is pre- that the present embodiment does not limit realization
Survey the side that the specific manifestation form of the intelligent electronic device of the method for user view, i.e. the present embodiment do not limit prediction user view
The hardware environment that method is applicable.
Each step in Fig. 2 is illustrated respectively below.
S200, the current characteristic information for obtaining user.
S210, the corresponding behavioural analysis of the user determined according to the user identity information in the current characteristic information of user
Model.
S220, the corresponding behavior of information input user by sign user's current state in the current characteristic information of user
In analysis model.
S230, the current behavior acquisition of information of user according to the corresponding Analysis model of network behaviors output of user and working as user
The related network data of preceding behavioural information.
S240, according to using network data user's is current to be determined to the result of the current behavior information of the user
Behavior.
Above-mentioned steps S200-S240 refers to the description in above-described embodiment one, is not repeated.
S250, the current intention according to the current behavior of the above-mentioned user determined prediction user.
Specifically, because the current behavior of user can reflect the wish of user, therefore, this implementation to a certain extent
Example can predict the current intention of user according to the current behavior and combination predetermined policy of user;As in the user determined
Current behavior is just in session, then the current intention that user can be predicted according to predetermined policy is possible for being not intended to by phone
Or immediate communication tool etc. is bothered, and then the present embodiment can make when user has call, using play busy tone or
The modes such as call voice assistant are gone to avoid being possibly realized bothering for user;For another example in the current behavior of the user determined
During on the road gone to the cinema, the current intention that user can be predicted according to predetermined policy is possible for wishing electricity of guarding the threshing floor
Shadow, and then the present embodiment can make it possible to recommend corresponding film information to user;For another example the user determined work as
Move ahead as when going shopping, can be pre- according to predetermined policy if current time has arrived lunch or date for dinner
The current intention for measuring user is possible for wishing to have a meal near market, and then the present embodiment can make to recommend surrounding to user
The restaurant information or snack information having higher rating are possibly realized.Herein not to the current of the user that is predicted according to predetermined policy
It is intended to carry out one by one for example, the present embodiment does not limit the specific manifestation form of the current intention of the user predicted, and this
Predetermined policy in embodiment can be set according to the actual requirements.
Embodiment three, the device for determining user's current behavior.
Fig. 3 is the schematic device of determination user's current behavior of the present embodiment.In the schematic device shown in Fig. 3,
The device of determination user's current behavior of the present embodiment mainly includes:Obtain characteristic information module 300, determine analysis model module
310th, message processing module 320, acquisition network data model 330 and determination current behavior module 340.Optionally, this implementation
The device of example also includes:Model setup module (not shown in Fig. 3).
Device described in the present embodiment can be arranged in intelligent electronic device, and the device is generally disposed at network side
Intelligent electronic device (being such as arranged at the server network equipment in high in the clouds) in, certain the present embodiment is also not excluded for the device and set
The possibility being placed in the intelligent electronic device of user side (such as intelligent mobile phone or tablet personal computer user equipment) is (wherein
Model can by network side the network equipment train, and be issued to user side).The present embodiment, which is not limited, is provided with determination user
The specific manifestation form of the intelligent electronic device of the device of current behavior, i.e. the present embodiment, which are not limited, determines user's current behavior
The hardware environment that device is applicable.Each module in Fig. 3 is described in detail respectively below.
Obtain the current characteristic information that characteristic information module 300 is mainly used in obtaining user.
Specifically, the current characteristic information of the user in the present embodiment is the user in current time or current slot
Characteristic information;The current characteristic information of user in the present embodiment can be stored as the history feature information of user;
That is, the current characteristic information of user and the history feature information of user are the characteristic information of user, in working as user
Preceding characteristic information lose its it is ageing when, the current characteristic information of user turns into the history feature information of user.
The characteristic information of user in the present embodiment refers to the information of the characteristic feature for describing user, is such as used to describe
The static nature of user and the information of behavioral characteristics.
The static nature of user in the present embodiment is properly termed as user property in user property, and the present embodiment can be with
Including:The sex of user, the age bracket (or year of birth generation of user) of user, the occupational group of user, the address of user, use
The business address at family, the incommutation instrument of user, the work and rest rule of user, the income level of user, the education water of user
Behavior preference of flat and user etc..
The behavioral characteristics of user in the present embodiment are properly termed as the state of user, and the state of the user in the present embodiment
It can specifically include:Environment that the motion state of user, user are in and the particular location where user etc..
The characteristic information of user in the present embodiment (i.e. believe by the current characteristic information of user and the history feature of user
Breath) generally include:The information of user's current state is characterized, and characterizes information sign user's current state of user's current state
Information can be specially sensor collection information, the satellite positioning information of user equipment in user equipment and based on user
Equipment microphone collection audio-frequency information in one or more.Sensor collection information in above-mentioned user equipment can be with
The information of gyroscope collection specially in user equipment, the information of the gravity sensor collection in user equipment and user set
One or more in the information of acceleration transducer collection in standby.The satellite positioning information of above-mentioned user equipment can have
Body is the location information based on GPS (G Iobal Position System, global positioning system) of user equipment, certainly, is used
The satellite positioning information of family equipment can also be specially location information based on big-dipper satellite alignment system of user equipment etc..On
The audio-frequency information for stating the microphone collection of user equipment be usually user sound in the environment.
The characteristic information of user in the present embodiment (i.e. believe by the current characteristic information of user and the history feature of user
Breath) generally also include:User's mark, the telephonic communication record in user equipment, the message registration of user, log information, Yong Hushe
For any one in each application message and user equipment model of middle installation or multiple.Above-mentioned user's mark can be specific
IMSI of Mobile Directory Number or user for user etc..The message registration of above-mentioned user is generally specially in pre- timing recently
Between in section user message registration.Above-mentioned log information can be specially user to be used using the historical record of each application, user
Inputted when the used network address and user carry out network access using browser when browser carries out network access
Search keyword etc..Each application message installed in above-mentioned user equipment can be specially the row for the application installed in user equipment
Table information.
Under normal conditions, the content that the characteristic information of the user of the present embodiment is included can comprehensive as far as possible one
A bit, so as to being conducive to accurately determining user's current behavior.In addition, it is necessary to special instruction, the spy of above-mentioned user
Reference ceases the content that the current characteristic information that included content is both user is included, and is also the history feature information institute of user
Comprising content.
In the case where the device of the present embodiment is arranged in user equipment, obtaining characteristic information module 300 can be with fixed
When or the mode of not timing obtain the current characteristic information of user, such as obtain the timing of characteristic information module 300 or not timing
The microphone of triggering user equipment carry out audio collection, with obtain the user in short time period sound in the environment;Again
Such as obtain the attitude information that the gyroscope in the collection user equipment of the timing of characteristic information module 300 or not timing is produced;
For another example, message registration and the day that user is once obtained before being obtained from of the timing of characteristic information module 300 or not timing are obtained
The message registration and log information of user produced by after will information time.Should be in addition, obtaining characteristic information module 300
In the case of the permission for obtaining user, the current characteristic information of user can be just got.
In the case where the device of the present embodiment is arranged in the network equipment of network side, user equipment can be using timing
Or the mode of not timing is obtained the current characteristic information of user and reported to network side, so that being arranged at the net of network side
Acquisition characteristic information module 300 in network equipment gets the current characteristic information of user.The present embodiment not restricting user equipment
To the specific implementation of the current characteristic information of network side report of user.
Determine that the user identity information that analysis model module 310 is mainly used in the current characteristic information according to user is determined
The corresponding Analysis model of network behaviors of user.
Specifically, the model setup module in the present embodiment would generally set respective behavior respectively for multiple users in advance
Multiple Analysis model of network behaviors, and each Analysis model of network behaviors pair are usually provided with the device of analysis model, i.e. the present embodiment
A user is answered, and different Analysis model of network behaviors generally corresponds to different users, i.e., different users has different behaviors
Analysis model.Model setup module can be by the storage corresponding with user identity information of each Analysis model of network behaviors, so as to obtain
Take after characteristic information module 300 gets the current characteristic information of user, determine that analysis model module 310 can be deposited in advance
The corresponding relation that the user identity information in the current characteristic information with user matches is searched in the information of storage, and will be matched
Corresponding relation in Analysis model of network behaviors be used as the corresponding Analysis model of network behaviors of the user.
Model setup module is mainly used in setting its corresponding Analysis model of network behaviors for user;And model setup module master
Including:First submodule, the second submodule and the 3rd submodule;Wherein, the first submodule is mainly used according to user's
History feature information determines the type belonging to the user property of the user;Second submodule is mainly used in many from what is pre-set
The basic model with the type matching belonging to user property is chosen in individual basic model;3rd submodule be mainly used according to
The history feature information at family sets personalized model parameter for the above-mentioned basic model matched, to form the corresponding behavior of user
Analysis model.
For a specific user, the model setup module of the present embodiment is believed using the history feature of the user
Cease and its corresponding Analysis model of network behaviors is set for the user, model setup module is that the user sets the one of Analysis model of network behaviors
Individual detailed process can be:
First, multiple basic models, and each basic model one kind of correspondence are previously provided with model setup module
Property type user, and the user of different basic model correspondence different attribute type, such as model setup module can be directed to court
This attribute type of working clan in nine evenings five sets a basic model, and for this using the working clan of flexible job system
Attribute type sets basic model etc.;In the example above, the classification of the user property type corresponding to basic model
Granularity is thicker, and in actual applications, the classification of the user property type corresponding to the basic model in model setup module
Granularity can be more careful, and the mode classification of the attribute type of user can be set according to the actual requirements.
Secondly, after acquisition characteristic information module 300 obtains the substantial amounts of history feature information of a user, model
Setup module (such as the first submodule) can be determined according to the substantial amounts of history feature information of the user attribute of the user with
And the type belonging to the attribute of the user, such as model setup module (such as the first submodule) is by by the substantial amounts of history of the user
In the user property disaggregated model that characteristic information input is pre-set, so that the present embodiment can be according to user property disaggregated model
The information of output determines the type belonging to the attribute of the user.User property disaggregated model in the present embodiment is by big
Formed by the history feature information of amount user is learnt, as the present embodiment can be using the mode of learning for having supervision to a large amount of
The history feature information of user is learnt, so as to form user property disaggregated model;The above-mentioned history feature to a large number of users
The result that information is learnt and the mode that user classifies are closely related.
Include in the user property of the present embodiment:The sex of user, the age bracket (or year of birth generation of user) of user,
The occupational group of user, the address of user, the business address of user, the incommutation instrument of user, the work and rest rule of user,
In the case of the Behavior preference of the income level of user, the level of education of user and user, user property disaggregated model can be with
According to log information (such as network access address and the search keyword of a large amount of (such as one month or some months) of a user
Deng) etc. determine that user likes the content browsed, and the satellite positioning information of substantial amounts of user equipment according to the user etc. is really
Place that the fixed user likes in leisure time etc., so as to be tentatively inferred to the sex of user, the age bracket of user with
And the level of education of user etc.;User property disaggregated model can be according to a large amount of (such as one month or some months) of a user
User equipment satellite positioning information determine user periodically in the residence time of some position, may thereby determine that out
The work and rest rule in the address of user, the business address of user and user etc., and the sensor collection information in user equipment
The auxiliary information for the work and rest rule for determining the address of user, the business address of user and user can be turned into;User property
Disaggregated model can according to a large amount of (such as one month or some months) telephone relations record (whether call frequent) of user,
The satellite positioning information (working time whether be commonly located at a small regional extent in etc.) and user equipment of user equipment
In the rough occupational group for determining user of sensor collection information etc., user property disaggregated model can exist according to user
The satellite positioning information of the user equipment of moving process can determine the friendship on and off duty of user between its address and business address
Logical instrument, and sensor collection information in user equipment can also turn into the auxiliary for determining the incommutation instrument of user
Information;The incommutation instrument of user and the address of user etc. also contribute to the further income level for determining user with
And the level of education of user etc.;User property disaggregated model can be according to the satellite fix of the substantial amounts of user equipment of a user
Information, the audio-frequency information of microphone collection based on user equipment and the work and rest rule of the above-mentioned user determined can be determined
Go out Behavior preference of user etc., restaurant that such as restaurant often gone at noon of user, user often go on day off, user are frequent
Shopping place that the frequency and time and user that the cinema that goes, user see a film often are gone etc., in addition, in user equipment
Sensor collection information can also turn into the auxiliary information for determining the Behavior preference of user.It above are only and user property is classified
Model determines one of the process of user property for example, the present embodiment, which does not limit user property disaggregated model, determines use
The specific implementation of family attribute.
As an example, the user property disaggregated model of the present embodiment can determine to use based on the partial content in user property
Type belonging to the attribute of family, such as user property disaggregated model can according to the sex of user, the age bracket of user, user occupation
Classification, the incommutation instrument of user, the income level of user and the level of education of user are classified to user property,
The present embodiment can pre-set multiple types, and the specific value of the parameter included according to each type according to the actual requirements
To determine the type belonging to the user property of active user, the present embodiment does not limit user property disaggregated model and determines user property
The specific implementation of affiliated type.
Again, model setup module (such as the second submodule) is after the type belonging to the user property is determined, can be with
The basic model with the type matching belonging to the user property is chosen from the multiple basic models pre-set;I.e. model is set
Put module (such as the second submodule) and be previously provided with multiple basic models, and each basic model is to that should have a kind of user to belong to
Property type, so as to determine and should adopt during type belonging to determining user property in model setup module (such as the second submodule)
Basic model.
Finally, model setup module (such as the 3rd submodule) can be user property according to the history feature information of the user
The basic model of affiliated type matching sets personalized model parameter, so as to form the corresponding behavioural analysis mould of the user
Type;As model setup module (such as the 3rd submodule) can using the user a large amount of history feature information to it is above-mentioned match
Basic model is trained, so that the model parameter of the basic model is the personalized model ginseng set for the user
Number, that is, the basic model after training is the corresponding Analysis model of network behaviors of user.Personalized model parameter in the present embodiment can be with
The model parameter for being the Behavior preference in the address based on user, the business address of user and user etc. and setting.
The Analysis model of network behaviors of the present embodiment can use based on+HMM that (including HMM and GMHMM etc. is using HMM as core
Algorithm) Analysis model of network behaviors, and identification sorting algorithm in the Analysis model of network behaviors based on+HMM can be using " forward direction is calculated
Method ".The present embodiment is no longer described in detail herein to the specific training process of Analysis model of network behaviors.
Special instruction is additionally needed, successfully its Analysis model of network behaviors has been provided with for user in model setup module
In the case of, the current characteristic information for obtaining the user that characteristic information module 300 is subsequently got can be by the history as user
Characteristic information is stored, and the history feature information of the user subsequently stored still can be used for model setup module to the user's
Analysis model of network behaviors is trained, with the personalized mould for the Analysis model of network behaviors for constantly correcting and (adjust or adjust) user
Shape parameter, enables the personalized model parameter of the Analysis model of network behaviors of user preferably to show actual conditions with user.
Message processing module 320 is mainly used in the information of sign user's current state in the current characteristic information by user
Input in the corresponding Analysis model of network behaviors of user;And the message processing module 320 includes:4th submodule, the 5th submodule,
Six submodules, the 7th submodule and the 8th submodule;4th submodule therein is mainly used in from current characteristic information set
It is middle to read untreated current characteristic information;5th submodule is mainly used in recognizing the class of each current characteristic information read out
Type;6th submodule is mainly used in choosing the information for characterizing user's current state according to the type identified;7th submodule master
Be used to the information of the different types of sign user current state of correspondence same time being combined, with formed using the time as
At least one status information group of dimension;8th submodule is mainly used in the corresponding behavioural analysis of status information group input user
In model.
Specifically, the information of sign user's current state in the current characteristic information of the user of the present embodiment can be specific
For sensor collection information, the satellite positioning information of user equipment in user equipment and Mike's elegance based on user equipment
The audio-frequency information of collection.
Message processing module 320 (such as the 4th submodule, the 5th submodule, the 6th submodule and the 7th submodule) can be with
First the information to sign user's current state in the current characteristic information of user is arranged, then, message processing module 320
(such as the 8th submodule) again inputs the information of sign user's current state after arrangement in the corresponding Analysis model of network behaviors of user.
Message processing module 320 is arranged to the information of sign user's current state in the current characteristic information of user
A specific example be:The current characteristic information got is stored in by the present embodiment in acquisition characteristic information module 300
In the case of in current characteristic information set, message processing module 320 (such as the 4th submodule) can be first from current characteristic information
The current characteristic information of the untreated user is read in set, message processing module 320 (such as the 5th submodule) identification is read
The type of each current characteristic information gone out, the type of current characteristic information can for address list type, (such as user sets posture type
Sensor collection information in standby belongs to posture type), audio types (believe by the audio of the microphone collection such as based on user equipment
Breath belongs to audio types) and align_type (satellite positioning information of such as user equipment belongs to align_type);Afterwards, information
Processing module 320 (such as the 6th submodule) can choose the information for characterizing user's current state, information according to the type identified
Processing module 320 (such as the 7th submodule) carries out the information of the different types of sign user current state of correspondence same time
Combination, so as to form at least one status information group by dimension of the time.
One specific example, obtains the sign in the current characteristic information for the user that characteristic information module 300 is got
The information of user's current state can be expressed as the form shown in above-mentioned formula (1), and message processing module 320 can be by above-mentioned public affairs
The finish message of sign user's current state in current characteristic information shown in formula (1) is the form shown in above-mentioned formula (2).
Message processing module 320 (such as the 8th submodule) can be incited somebody to action according to the sequencing (i.e. time sequencing) of time dimension
Each status information group is sequentially input in the corresponding Analysis model of network behaviors of user.Such as message processing module 320 (such as the 8th submodule)
First (Motion Type at tn-1, Sound Type at tn-1, Location Type at tn-1) can be inputted and used
In the corresponding Analysis model of network behaviors in family, then, message processing module 320 (such as the 8th submodule) is by each between tn-1 and tn
Between the status information group at moment sequentially input in the corresponding Analysis model of network behaviors of user, until will (Motion Type at tn,
Sound Type at tn, Location Type at tn).
Acquisition network data model 330 is mainly used in the current of the user exported according to the corresponding Analysis model of network behaviors of user
Behavioural information obtains the network data related to the current behavior information of user.Obtaining network data model 330 can include:The
Nine submodules and the tenth submodule;9th submodule therein is mainly used in Analysis model of network behaviors output corresponding from the user
User current behavior information in extract keyword;Tenth submodule therein is mainly used in searching and keyword from network
Related network data.
Specifically, the current behavior information of the user in the present embodiment is the information for describing the current behavior of user,
And the current behavior of user is referred to as the current life-form structure of user or the event of user etc..The current behavior of user can
Be specially over/under on the road of class (using/driving/public transport mode of walking), removing certain destination (such as film
Institute) road on (using/driving/public transport mode of walking), having lunch (at the restaurant/family/office), in work
(/ bending over one's desk working/on the road gone on business just in session), taking a nap or going shopping etc..The present embodiment is not limited
The specific manifestation form of user's current behavior processed.
The Analysis model of network behaviors of user in the present embodiment can according to input sign user's current state information (such as
Sensor collection information, the satellite positioning information of user equipment in user equipment and the microphone collection based on user equipment
Audio-frequency information) obtain the probability of each possible user behavior, the Analysis model of network behaviors of such as user may for each
User behavior be utilized respectively pre-defined algorithm to input sign user's current state information carry out probability calculation, it is defeated to obtain
The probability of each possible user behavior of the information correspondence of the sign user's current state entered;The Analysis model of network behaviors of user exists
Calculating during the probability of each possible user behavior would generally go through to characterizing the information of user's current state, user
The many-side such as historical event part and log information carries out comprehensive consideration.The present embodiment does not limit the Analysis model of network behaviors of user to input
Sign user's current state information calculated implement process.
The present embodiment obtains network data model 330 in order that the current behavior of the user finally determined is more accurate
The network data of correlation can be obtained, with the use exported by determination current behavior module 340 to the Analysis model of network behaviors of user
The current behavior information at family is verified, the Analysis model of network behaviors institute with user can be obtained as obtained network data model 330
The related network data of the current behavior information of the probability highest user of output, with by 340 pairs of determination current behavior module with
The current behavior information for the probability highest user that the Analysis model of network behaviors at family is exported is verified, network data is obtained for another example
Module 330 can obtain the probability highest exported with the Analysis model of network behaviors of user and the current behavior letter of secondary high user
The related network data of breath difference, with the probability exported by determination current behavior module 340 for the Analysis model of network behaviors of user
Highest and the current behavior information of secondary high user are verified respectively.
Obtain network data model 330 and obtain the specific example of network data of correlation and be:Obtain network data
The current behavior information for the probability highest user that module 330 (such as the 9th submodule) is exported for the Analysis model of network behaviors of user
Word segmentation processing is carried out, then the keyword in current behavior information to extract user, obtains network data model 330 (such as the
Ten submodules) keyword that is extracted using the 9th submodule searches the network data related to keyword from network, and such as the
Ten submodules can search for the network data related to keyword using modes such as web crawlers.The present embodiment does not limit the 9th son
The specific implementation of the current behavior information extraction keyword for the user that module is exported from the Analysis model of network behaviors of user with
And the tenth submodule the specific implementation of the network data related to keyword is searched from network.
Determine current behavior module 340 be mainly used according to using above-mentioned network data to the current behavior information of user
The result determines the current behavior of user.
Specifically, determining that current behavior module 340 can be according to the above-mentioned related network data got to user's
Current behavior information (the current behavior letter of such as probability highest user of the user for the corresponding probability that Analysis model of network behaviors is exported
Breath) verified, for example, determining that current behavior module 340 verifies the current behavior information of probability highest user whether with obtaining
The corresponding network data got is consistent, if be consistent, it is determined that current behavior module 340 determines to be verified, so that it is determined that
The current behavior information for the probability highest user that current behavior module 340 can be exported the Analysis model of network behaviors of user is made
For the information for the current behavior for describing user;If be not consistent, it is determined that current behavior module 340 determines authentication failed, so that
Determine the current behavior letter for the probability highest user that current behavior module 340 can not be exported the Analysis model of network behaviors of user
Cease the information of the current behavior as description user.Above-mentioned to be consistent can be specially that the time parameter such as place is consistent, a tool
The example of body, the current behavior information of the probability highest user exported in the Analysis model of network behaviors of user represents that user is going
In the case of on toward the road in the first gymnasium, if obtain network data model 330 using the first gymnasium acquired in keyword
To network data include certain singer and in the first gymnasium will hold concert when the A B month C day D, it is determined that current behavior
Module 340 is determining the current date for A B months C day, and current time is closed in the case of D, determines current behavior mould
Block 340 can determine that this is verified, so that the determination current behavior module 340 of the present embodiment can accurately determine use
The current behavior at family is is going on the road in the first gymnasium, conversely, determining that current behavior module 340 can not accurately then be determined
Whether the current behavior of user is to go on the road in the first gymnasium.
Determine that the current behavior for the user that current behavior module 340 is determined can be not only used for predicting user view, and
And other scenes are can be used for, it can such as remember by using the current behavior for determining the user that current behavior module 340 is obtained
Life course of the family in the range of a period of time is employed, so as to be presented to user when user checks or regularly present
To user etc..The present embodiment does not limit the concrete application for determining the current behavior for the user that current behavior module 340 is determined
Scape.
Example IV, the device for predicting user view.
Fig. 4 is the schematic device of the prediction user view of the present embodiment.In the schematic device shown in Fig. 4, this reality
Applying the device of the prediction user view of example includes:Obtain characteristic information module 300, determine analysis model module 310, information processing
Module 320, obtain network data model 330, determine current behavior module 340 and Intention Anticipation module 450.Optionally, originally
The device of embodiment also includes:Model setup module (not shown in Fig. 4).
Device described in the present embodiment can be arranged in intelligent electronic device, and the device is generally disposed at network side
Intelligent electronic device (being such as arranged at the server network equipment in high in the clouds) in, certain the present embodiment is also not excluded for the device and set
The possibility being placed in the intelligent electronic device of user side (such as intelligent mobile phone or tablet personal computer user equipment) is (wherein
Model can by network side the network equipment train, and be issued to user side).The present embodiment, which is not limited, is provided with prediction user
The device that the specific manifestation form of the intelligent electronic device of the device of intention, i.e. the present embodiment do not limit prediction user view is fitted
Hardware environment.
Characteristic information module 300 is obtained, analysis model module 310, message processing module 320 is determined, obtains network data
The description in operation such as above-mentioned embodiment three performed by module 330 and determination current behavior module 340, is not repeated herein
Illustrate, the Intention Anticipation module 450 in Fig. 4 is described in detail below.
Intention Anticipation module 450 is mainly used according to the current behavior for determining the user that current behavior module 340 is determined
Predict the current intention of user.
Specifically, because the current behavior of user can reflect the wish of user to a certain extent, therefore, it is intended that in advance
The current intention of user can be predicted according to the current behavior and combination predetermined policy of user by surveying module 450;Such as it is determined that working as
The current behavior for the user that preceding behavioral module 340 is determined is just in session, then Intention Anticipation module 450 can be according to predetermined plan
The current intention for slightly predicting user is possible for being not intended to being bothered by phone or immediate communication tool etc., and then the present embodiment
Device can make when user has call, avoided pair using playing busy tone or going to the modes such as call voice assistant
Bothering for user is possibly realized;For another example it is determined that the current behavior for the user that current behavior module 340 is determined is to remove film
When on the road of institute, it is intended that the current intention that prediction module 450 can predict user according to predetermined policy is possible for wishing
Guard the threshing floor film, and then the device of the present embodiment can make it possible to recommend corresponding film information to user;For another example it is determined that
The current behavior for the user that current behavior module 340 is determined be when going shopping, if current time arrived lunch or
Person's date for dinner, then Intention Anticipation module 450 can be predicted according to predetermined policy user current intention be possible for wish
Had a meal near market, and then the device of the present embodiment can make the restaurant information or small for recommending surrounding to have higher rating to user
Information is eaten to be possibly realized.The current intention for the user not predicted herein to Intention Anticipation module 450 according to predetermined policy is carried out
One by one for example, the present embodiment does not limit the specific manifestation shape of the current intention for the user that Intention Anticipation module 450 is predicted
Formula, and the predetermined policy that Intention Anticipation module 450 is used can set according to the actual requirements.
It should be noted that a part of of the present invention can be applied to computer program product, such as computer program
Instruction, when it is performed by intelligent electronic device (such as intelligent mobile phone or tablet personal computer), is set by the smart electronicses
The method according to the invention and/or technical scheme can be called or provided to standby operation.And call the journey of the method for the present invention
Sequence is instructed, and is possibly stored in fixed or moveable recording medium, and/or passes through broadcast or other signaling bearers matchmaker
Data flow in body and be transmitted, and/or be stored in and deposited according to the work of the intelligent electronic device of described program instruction operation
In reservoir.Here, including a device according to one embodiment of present invention, the device includes referring to for storing computer program
The memory of order and the processor for execute program instructions, wherein, when the computer program instructions are by the computing device,
Trigger methods and/or techniques scheme of the plant running based on foregoing multiple embodiments according to the present invention.
It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie
In the case of without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, embodiment all should be regarded as exemplary, and be nonrestrictive, the scope of the present invention is by appended power
Profit is required rather than described above is limited, it is intended that all in the implication and scope of the equivalency of claim by falling
Change is included in the present invention.Any reference in claim should not be considered as to the claim involved by limitation.This
Outside, it is clear that the word of " comprising " one is not excluded for other units or step, and odd number is not excluded for plural number.That is stated in system claims is multiple
Unit or device can also be realized by a unit or device by software or hardware.The first, the second grade word is used for table
Show title, and be not offered as any specific order.
Claims (10)
1. a kind of method for determining user's current behavior, wherein, it the described method comprises the following steps:
Obtain the current characteristic information of user;
User identity information in the current characteristic information of the user determines the corresponding Analysis model of network behaviors of the user;
The information of sign user's current state in the current characteristic information of the user is inputted into the corresponding behavior of the user
In analysis model;
The current behavior acquisition of information of the user exported according to the corresponding Analysis model of network behaviors of the user is worked as with the user's
The related network data of preceding behavioural information;
The user's is current to be determined to the result of the current behavior information of the user according to using the network data
Behavior;
Wherein, the current characteristic information of the user is stored as the history feature information of the user, the history feature letter
Cease the model parameter for setting and updating the corresponding Analysis model of network behaviors of the user.
2. according to the method described in claim 1, wherein, the setting up procedure of the corresponding Analysis model of network behaviors of the user includes:
Type according to belonging to the history feature information of the user determines the user property of the user;
The basic model with the type matching belonging to the user property is chosen from the multiple basic models pre-set;
Personalized model parameter is set for the basic model matched according to the history feature information of the user, to be formed
The corresponding Analysis model of network behaviors of the user.
3. method according to claim 2, wherein, the history feature information according to the user determines the user
Type belonging to attribute includes:
By in the history feature information input user property disaggregated model of the user;
The user property disaggregated model determines user property and described according to the history feature information of the user of input
Type belonging to user property, and export the type information.
4. method according to claim 2, wherein, the user property includes:The sex of user, the age bracket of user,
The occupational group of user, the address of user, the business address of user, the incommutation instrument of user, the work and rest rule of user,
The Behavior preference of the income level of user, the level of education of user and user.
5. according to the method described in claim 1, wherein, it is described characterize user's current state information include:In user equipment
Sensor collection information, the satellite positioning information of user equipment and based on user equipment microphone collection audio-frequency information
In at least one.
6. method according to claim 5, wherein, the current characteristic information of the user also includes:In user equipment
Each application message and user equipment model installed in telephonic communication record, the message registration of user, log information, user equipment
In any one or it is multiple.
7. the method according to any claim in claim 1 to 6, wherein, the current signature by the user
The information of sign user's current state in information, which inputs the corresponding Analysis model of network behaviors of the user, to be included:
Untreated current characteristic information is read from current characteristic information set;
The type of each current characteristic information read out described in identification;
The information for characterizing user's current state is chosen according to the type identified;
The information of the different types of sign user current state of correspondence same time is combined, to be formed using the time as dimension
At least one status information group of degree;
The status information group is inputted in the corresponding Analysis model of network behaviors of the user.
8. the method according to any claim in claim 1 to 6, wherein, it is described according to the corresponding row of the user
Include for the acquisition of information of the analysis model output network data related to the information of the output:
Keyword is extracted from the current behavior information of the user of the corresponding Analysis model of network behaviors output of the user;
The network data related to the keyword is searched from network.
9. a kind of method for predicting user view, wherein, methods described includes:
Obtain the current characteristic information of user;
User identity information in the current characteristic information of the user determines the corresponding Analysis model of network behaviors of the user;
The information of sign user's current state in the current characteristic information of the user is inputted into the corresponding behavior of the user
In analysis model;
The current behavior acquisition of information of the user exported according to the corresponding Analysis model of network behaviors of the user is worked as with the user's
The related network data of preceding behavioural information;
The user's is current to be determined to the result of the current behavior information of the user according to using the network data
Behavior;
The current intention of the user is predicted according to the current behavior of the user;
Wherein, the current characteristic information of the user is stored as the history feature information of the user, the history feature letter
Cease the model parameter for setting and updating the corresponding Analysis model of network behaviors of the user.
10. a kind of device for determining user's current behavior, wherein, described device includes:
Obtain characteristic information module, the current characteristic information for obtaining user;
Analysis model module is determined, the use is determined for the user identity information in the current characteristic information according to the user
The corresponding Analysis model of network behaviors in family;
Message processing module, for the information of sign user's current state in the current characteristic information of the user to be inputted into institute
State in the corresponding Analysis model of network behaviors of user;
Network data model is obtained, the current behavior of the user for being exported according to the corresponding Analysis model of network behaviors of the user is believed
Breath obtains the network data related to the current behavior information of the user;
Current behavior module is determined, for according to checking knot of the utilization network data to the current behavior information of the user
Fruit determines the current behavior of the user;
Wherein, the current characteristic information of the user is stored as the history feature information of the user, the history feature letter
Cease the model parameter for setting and updating the corresponding Analysis model of network behaviors of the user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710396359.9A CN107194515A (en) | 2017-05-27 | 2017-05-27 | Determine user's current behavior and the method and apparatus for predicting user view |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710396359.9A CN107194515A (en) | 2017-05-27 | 2017-05-27 | Determine user's current behavior and the method and apparatus for predicting user view |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107194515A true CN107194515A (en) | 2017-09-22 |
Family
ID=59876181
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710396359.9A Pending CN107194515A (en) | 2017-05-27 | 2017-05-27 | Determine user's current behavior and the method and apparatus for predicting user view |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107194515A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107798341A (en) * | 2017-09-30 | 2018-03-13 | 平安科技(深圳)有限公司 | User view Forecasting Methodology, electronic equipment and computer-readable recording medium |
CN108124197A (en) * | 2017-12-18 | 2018-06-05 | 广东省电信规划设计院有限公司 | The recognition methods of terminal access behavior and device |
CN109325845A (en) * | 2018-08-15 | 2019-02-12 | 深圳市和讯华谷信息技术有限公司 | A kind of financial product intelligent recommendation method and system |
CN110619406A (en) * | 2018-06-19 | 2019-12-27 | 中移信息技术有限公司 | Method and device for determining business abnormity |
CN114158166A (en) * | 2020-09-07 | 2022-03-08 | 中国联合网络通信集团有限公司 | Control method and device of lighting equipment |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101887489A (en) * | 2010-05-24 | 2010-11-17 | 陈益强 | Method for interactive influence of characters in real world and virtual world |
CN103390194A (en) * | 2012-05-07 | 2013-11-13 | 北京三星通信技术研究有限公司 | Method, device and system for predicating user intention and recommending suggestion |
CN104679743A (en) * | 2013-11-26 | 2015-06-03 | 阿里巴巴集团控股有限公司 | Method and device for determining preference model of user |
CN105069077A (en) * | 2015-07-31 | 2015-11-18 | 百度在线网络技术(北京)有限公司 | Search method and device |
CN105354196A (en) * | 2014-08-19 | 2016-02-24 | 北京羽扇智信息科技有限公司 | Information pushing method and information pushing apparatus |
CN105677780A (en) * | 2014-12-31 | 2016-06-15 | Tcl集团股份有限公司 | Scalable user intent mining method and system thereof |
CN107436897A (en) * | 2016-05-26 | 2017-12-05 | 冯小平 | A kind of determination method and device of user's current behavior |
-
2017
- 2017-05-27 CN CN201710396359.9A patent/CN107194515A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101887489A (en) * | 2010-05-24 | 2010-11-17 | 陈益强 | Method for interactive influence of characters in real world and virtual world |
CN103390194A (en) * | 2012-05-07 | 2013-11-13 | 北京三星通信技术研究有限公司 | Method, device and system for predicating user intention and recommending suggestion |
CN104679743A (en) * | 2013-11-26 | 2015-06-03 | 阿里巴巴集团控股有限公司 | Method and device for determining preference model of user |
CN105354196A (en) * | 2014-08-19 | 2016-02-24 | 北京羽扇智信息科技有限公司 | Information pushing method and information pushing apparatus |
CN105677780A (en) * | 2014-12-31 | 2016-06-15 | Tcl集团股份有限公司 | Scalable user intent mining method and system thereof |
CN105069077A (en) * | 2015-07-31 | 2015-11-18 | 百度在线网络技术(北京)有限公司 | Search method and device |
CN107436897A (en) * | 2016-05-26 | 2017-12-05 | 冯小平 | A kind of determination method and device of user's current behavior |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107798341A (en) * | 2017-09-30 | 2018-03-13 | 平安科技(深圳)有限公司 | User view Forecasting Methodology, electronic equipment and computer-readable recording medium |
WO2019061990A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技(深圳)有限公司 | User intention prediction method, electronic device, and computer readable storage medium |
CN108124197A (en) * | 2017-12-18 | 2018-06-05 | 广东省电信规划设计院有限公司 | The recognition methods of terminal access behavior and device |
CN108124197B (en) * | 2017-12-18 | 2020-09-18 | 广东省电信规划设计院有限公司 | Method and device for identifying terminal access behavior |
CN110619406A (en) * | 2018-06-19 | 2019-12-27 | 中移信息技术有限公司 | Method and device for determining business abnormity |
CN109325845A (en) * | 2018-08-15 | 2019-02-12 | 深圳市和讯华谷信息技术有限公司 | A kind of financial product intelligent recommendation method and system |
CN114158166A (en) * | 2020-09-07 | 2022-03-08 | 中国联合网络通信集团有限公司 | Control method and device of lighting equipment |
CN114158166B (en) * | 2020-09-07 | 2023-05-16 | 中国联合网络通信集团有限公司 | Control method and device of lighting equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107302566A (en) | The method and apparatus of pushed information | |
CN107194515A (en) | Determine user's current behavior and the method and apparatus for predicting user view | |
CN109034660B (en) | Method and related device for determining risk control strategy based on prediction model | |
Liu et al. | Extracting semantic location from outdoor positioning systems | |
CN104246748B (en) | System and method for determining situation | |
US20160105801A1 (en) | Geo-based analysis for detecting abnormal logins | |
CN107977734B (en) | Prediction method based on mobile Markov model under space-time big data | |
CN105894372A (en) | Method and device for predicting group credit | |
CN106778876A (en) | User classification method and system based on mobile subscriber track similitude | |
CN112084413B (en) | Information recommendation method, device and storage medium | |
CN103530304A (en) | On-line recommendation method, system and mobile terminal based on self-adaption distributed computation | |
CN105809557A (en) | Method and device for mining genders of users in social network | |
CN109117856A (en) | People and object method for tracing, apparatus and system based on intelligent edge cloud | |
CN115408621B (en) | Interest point recommendation method considering auxiliary information characteristic linear and nonlinear interaction | |
CN110191183A (en) | Accurate intelligent method for pushing, system, device and computer readable storage medium | |
CN109376901A (en) | A kind of service quality prediction technique based on decentralization matrix decomposition | |
CN110008402A (en) | A kind of point of interest recommended method of the decentralization matrix decomposition based on social networks | |
CN103399900A (en) | Image recommending method based on location service | |
CN107274217A (en) | Determine user's current behavior and the method and apparatus for predicting user view | |
CN107368499A (en) | A kind of client's tag modeling and recommendation method and device | |
CN110309339A (en) | Picture tag generation method and device, terminal and storage medium | |
CN108648017B (en) | User requirement matching method, device, equipment and storage medium easy to expand | |
CN107436897A (en) | A kind of determination method and device of user's current behavior | |
CN111353001B (en) | Method and device for classifying users | |
CN107832848A (en) | application management method, device, storage medium and electronic equipment |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170922 |