CN107704586B - User portrait method, device and system based on user activity address - Google Patents
User portrait method, device and system based on user activity address Download PDFInfo
- Publication number
- CN107704586B CN107704586B CN201710930889.7A CN201710930889A CN107704586B CN 107704586 B CN107704586 B CN 107704586B CN 201710930889 A CN201710930889 A CN 201710930889A CN 107704586 B CN107704586 B CN 107704586B
- Authority
- CN
- China
- Prior art keywords
- user
- address
- physical geographic
- user activity
- label
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/535—Tracking the activity of the user
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Fuzzy Systems (AREA)
- Computer Hardware Design (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a user portrait method based on a user activity address and a corresponding system, the method analyzes the meaning of the physical geographic position or/and the IP address used by the user in different time periods relative to the user activity address by acquiring and analyzing the physical geographic position or/and the IP address of the user, obtains the final user portrait result based on the user activity address by analyzing the time period and the corresponding frequency of the user using the physical geographic position or/and the IP address, and finally further inputs the user portrait result into an artificial neural network training or reasoning machine, thereby automatically constructing a more accurate user portrait.
Description
Technical Field
The invention relates to the field of data communication, in particular to a user portrait method, device and system based on a user activity address.
Background
A mobile terminal user representation is typically a collection of tags characterizing a mobile terminal user. The label set comprises sub characteristic label sets of a home address, a daily address, a learning unit, a working unit, a frequent dining place, a leisure and entertainment place, a tourist site and the like of a mobile terminal user.
At present, the method for collecting the mobile terminal user portrait sub-feature tag set mainly includes manual filling, obtaining according to the online behaviors of a user such as a webpage browsing record and an internet surfing trace, and the like. These methods have problems of "troublesome operation and insufficient data accuracy".
In view of the above, the present invention provides a method for user profiling based on a user activity address. The sub-characteristic label set of the user portrait of the user such as the home address, the daily address, the learning unit, the working unit, the frequently-going dining place, the leisure entertainment place, the tourist site and the like of the user is calculated through the activity frequency of the activity address of the mobile terminal user in a certain period of time, so that the operation cost of data acquisition is reduced, and the data accuracy is improved.
Disclosure of Invention
The invention aims to provide a user portrait method based on a user activity address, which can be used for automatically constructing a more accurate and fine user portrait.
The technical scheme adopted by the invention is as follows:
in one aspect, the present invention provides a method for user representation based on a user activity address, comprising:
step 1:
(1.1) presetting a user portrait label library of user activity addresses based on a time interval label and a frequency threshold, wherein the user activity addresses refer to a plurality of similar physical geographic positions or/and IP addresses in the same or within a set distance threshold;
(1.2) establishing an artificial neural network in advance through candidate user image data samples based on user activity addresses, and training the neural network by using a learning training module until the network converges; or establishing a knowledge base through a candidate user portrait data sample based on a user activity address, wherein the knowledge base is a knowledge map;
step 2: presetting the close physical geographic position or/and the IP address distance threshold;
and step 3:
(3.1) calculating time interval labels and frequency data of a plurality of similar physical geographic positions or/and IP addresses of the user in the same or within a set distance threshold;
the specific implementation process of the step (3.1) is as follows: acquiring a physical geographic position or/and an IP address when a user surfs the internet by using a mobile terminal, acquiring a corresponding time interval label according to a time interval when the user uses the same or a plurality of similar physical geographic positions or/and IP addresses within a set distance threshold, and counting the frequency of using the corresponding physical geographic position or/and IP address by the user within the time interval;
(3.2) matching a preset user portrait label library of the user activity address based on the time interval label and the frequency threshold, and calculating a user portrait data candidate based on the user activity address; if 1 candidate is obtained, it is adopted as the user portrait result. If multiple candidates are obtained, entering step 4;
the specific implementation process of the step (3.2) is as follows: matching the user portrait label library of the user activity address based on the time period label and the frequency threshold value preset in the step (1.1) according to the time period label and the frequency data obtained in the step (3.1), taking the corresponding label in the user portrait label library as a user portrait label candidate based on the user activity address, and when the user uses different physical geographic positions or/and IP addresses in the same time period, obtaining a plurality of user portrait label candidates based on the user activity address, and combining the user portrait label candidate based on the user activity address, the corresponding user activity address, the time period label of the user activity address used by the user, the time length of one-time use, the use frequency, the number of times and the accumulated time length in the time period into a user portrait data candidate based on the user activity address;
and 4, step 4: inputting the user portrait data candidate item based on the user activity address obtained in the step (3) into the artificial neural network trained in the step (1.2) to obtain an optimal user portrait result based on the user activity address; or matching corresponding rules from the knowledge base through an inference engine until matching is successful, and obtaining an optimal user portrait result based on a user activity address;
the specific implementation process of the step 4 is as follows: inputting the user portrait data candidate item processed in the step (3) based on the user activity address into an artificial neural network, wherein an input layer of the artificial neural network recognizes a label candidate item of the user portrait based on the user activity address, and a corresponding user activity address, a time period label of the user activity address used by the user, a time length of one-time use, a frequency of use in a time period, a frequency of use in the time period, and an accumulated time length as input parameters, and the input parameters are transmitted to an output layer through a hidden layer, so that an optimal user portrait result based on the user activity address is obtained by the output layer; or matching corresponding rules from the knowledge base through an inference engine until matching is successful, and obtaining an optimal user portrait result based on a user activity address;
and 5: presetting a time span, monitoring the change of a plurality of similar physical geographic positions or/and IP address data which are repeatedly used by a user for a plurality of times in a fixed time interval and are in the same or in a set distance threshold, and automatically adjusting the user portrait result based on the user activity address according to the data change and the processes of the steps 3 and 4.
Further, a method of user portrayal based on user activity address, the slot tag comprising: work hours, dining hours, sleeping hours, leisure hours, weekend hours, holidays, roaming to a foreign place; the user profile tag includes: companies, institutions, schools, home addresses, dormitories, hotels; the time interval labels are preset by a system and can be adjusted by a user according to the actual conditions of the user, and each time interval label sets the corresponding time interval length.
Further, a method for user representation based on user activity address, the method for the close physical geographic location or/and IP address distance threshold in step 2 can be preset by the system, and can also be manually set by the user.
Further, a method for user representation based on user activity address, the method for obtaining physical geographic location or/and IP address when user accesses internet by using mobile terminal, includes:
judging the internet access mode of a user mobile terminal, if a user uses a mobile communication network to access the internet, acquiring the communication base station information accessed by the mobile phone of the user, and acquiring the physical geographic position corresponding to the communication base station;
if the user uses WiFi to surf the internet, acquiring an IP address corresponding to the WiFi used by the user, and acquiring a physical geographic position corresponding to the IP address;
and if the user allows to use the GPS positioning, directly acquiring the corresponding physical geographic position according to the GPS positioning.
Further, a method for user representation based on user activity address, after obtaining the physical geographic location or/and IP address used by the user, querying a corresponding precise place name, includes: and obtaining an accurate place corresponding to the physical geographic position or/and the IP address through network search according to the obtained physical geographic position or/and the IP address, and obtaining the accurate place name.
In a second aspect, the present invention provides an apparatus for user representation based on a user activity address, comprising:
the device comprises a physical geographic position or/and IP address acquisition device, a server and a server, wherein the physical geographic position or/and IP address acquisition device is used for acquiring the physical geographic position or/and IP address when a user uses the mobile terminal to surf the internet, and corresponding information of use time and use time length and uploading the information to the server;
the device comprises a distance threshold setting device for setting and judging a plurality of close physical geographic positions or/and IP addresses to be a distance threshold corresponding to a user activity address according to the actual situation of a user;
and the user portrait result receiving device is used for receiving the user portrait result which is obtained by the server after calculation and is based on the user activity address, and providing further personalized service for the user according to the user portrait result based on the user activity address.
In a third aspect, the present invention provides a system for user representation based on a user activity address, comprising:
a mobile terminal;
a server, the server comprising: a user physical geographic position or/and IP address storage unit, which is used for storing the physical geographic position or/and IP address when the user registered on the server uses the corresponding mobile terminal, and the time information when the user uses the IP address or/and is at the physical geographic position;
the user portrait label library unit is used for storing time periods and frequency of user activity addresses corresponding to a plurality of similar physical geographic positions or/and IP addresses of a user in the same or within a set distance threshold value, and judging the corresponding relation between the time period labels, the frequency threshold value and the user activity addresses and the user portrait labels;
the user portrait construction unit is used for obtaining a user portrait based on a user activity address by inference according to a user physical geographic position or/and an IP address, and comprises the following steps: the system comprises a module for calling a physical geographic position or/and an IP address of a user, a time period analysis device for analyzing use time and use time length information corresponding to the physical geographic position or/and the IP address when the user uses a mobile terminal to surf the internet, a frequency calculation device for calculating the user activity address repeatedly used by the user for multiple times in a fixed time period, and an artificial neural network training model or a user portrait inference device consisting of a knowledge base and an inference machine;
and the user portrait output device is used for outputting the user portrait based on the user activity address obtained by the user portrait construction unit.
Further, a system for user representation based on a user activity address, comprising:
the apparatus for user representation based on user activity address further comprises: the time length presetting device is used for customizing a user data monitoring time length by a user so as to monitor the change of a physical geographic position or/and an IP address;
the server further comprises:
the user portrait adjusting unit is used for monitoring the data change condition of the user using the physical geographic position or/and the IP address in a fixed time period within a preset time length, and if the data change exceeds a preset threshold value, adjusting the user portrait based on the user activity address;
the mobile terminal network judging device is used for analyzing the internet access mode of the mobile terminal connected to the server;
the physical geographic position information determining device is used for acquiring a corresponding physical geographic position or/and an IP address according to the internet access mode of the mobile terminal and finally acquiring physical geographic position information;
and the accurate place network searching device is used for determining the accurate place name corresponding to the physical geographic position or/and the IP address through network searching.
The invention has the beneficial effects that: the invention provides a user portrait method based on a user activity address and a corresponding system, the method analyzes the meaning of the physical geographic position or/and the IP address used by a user in different periods relative to the user activity address by acquiring and analyzing the physical geographic position or/and the IP address of the user, obtains the final user portrait result based on the user activity address by analyzing the period and the corresponding frequency of the user using the physical geographic position or/and the IP address and finally further inputting artificial neural network training, thereby automatically constructing a more accurate user portrait.
Drawings
FIG. 1 is a flow diagram of a method for user profiling based on a user activity address;
FIG. 2 is a schematic diagram of a system for user representation based on user activity addresses;
FIG. 3 is a schematic diagram of a user representation building unit in FIG. 2.
Detailed Description
The invention is further illustrated by the following figures and examples.
As shown in FIG. 1, a method of user profiling based on a user activity address, comprising:
step 1:
(1.1) presetting a user portrait label library of user activity addresses based on a time interval label and a frequency threshold, wherein the user activity addresses refer to a plurality of similar physical geographic positions or/and IP addresses in the same or within a set distance threshold;
(1.2) establishing an artificial neural network in advance through candidate user image data samples based on user activity addresses, and training the neural network by using a learning training module until the network converges; or establishing a knowledge base through a candidate user portrait data sample based on a user activity address, wherein the knowledge base is a knowledge map;
step 2: presetting the close physical geographic position or/and the IP address distance threshold;
and step 3:
(3.1) calculating time interval labels and frequency data of a plurality of similar physical geographic positions or/and IP addresses of the user in the same or within a set distance threshold;
the specific implementation process of the step (3.1) is as follows: acquiring a physical geographic position or/and an IP address when a user surfs the internet by using a mobile terminal, acquiring a corresponding time interval label according to a time interval when the user uses the same or a plurality of similar physical geographic positions or/and IP addresses within a set distance threshold, and counting the frequency of using the corresponding physical geographic position or/and IP address by the user within the time interval;
the specific implementation process of the step (3.2) is as follows: matching the user portrait label library of the user activity address based on the time period label and the frequency threshold value preset in the step (1.1) according to the time period label and the frequency data obtained in the step (3.1), taking the corresponding label in the user portrait label library as a user portrait label candidate based on the user activity address, and when the user uses different physical geographic positions or/and IP addresses in the same time period, obtaining a plurality of user portrait label candidates based on the user activity address, and carrying out datamation on the user portrait label candidate based on the user activity address, the corresponding user activity address, the time period label of the user activity address used by the user, the time length of one-time use, the use frequency and the accumulated time length in the time period to form a user portrait data candidate based on the user activity address;
(3.2) matching a preset user portrait label library of the user activity address based on the time interval label and the frequency threshold, and calculating a user portrait data candidate based on the user activity address; if 1 candidate is obtained, the candidate is adopted as a user portrait result; if multiple candidates are obtained, entering step 4;
the specific implementation process of the step (3.2) is as follows: matching the user portrait label library of the user activity address based on the time period label and the frequency threshold value preset in the step (1.1) according to the time period label and the frequency data obtained in the step (3.1), taking the corresponding label in the user portrait label library as a user portrait label candidate based on the user activity address, and when the user uses different physical geographic positions or/and IP addresses in the same time period, obtaining a plurality of user portrait label candidates based on the user activity address, and combining the user portrait label candidate based on the user activity address, the corresponding user activity address, the time period label of the user activity address used by the user, the time length of one-time use, the use frequency, the number of times and the accumulated time length in the time period into a user portrait data candidate based on the user activity address;
and 4, step 4: inputting the user portrait data candidate item based on the user activity address obtained in the step (3) into the artificial neural network trained in the step (1.2) to obtain an optimal user portrait result based on the user activity address; or matching corresponding rules from the knowledge base through an inference engine until matching is successful, and obtaining an optimal user portrait result based on a user activity address;
the embodiment is suitable for accurately constructing the user portrait, and the influence of the change of the information on the user portrait is distinguished through the training of the artificial neural network by using a plurality of groups of physical geographic positions or/and IP addresses used by the user and the corresponding information, so that the accuracy of the user portrait can be effectively improved.
The invention constructs the user portrait according to the user activity address where the user often locates in a specific time period, and simultaneously, in order to ensure that the user portrait result is as reasonable and accurate as possible, the following core technical steps are adopted: (1) presetting a distance threshold value of similar physical geographic positions or/and IP addresses, and taking the address of a user doing activities in the same or a plurality of similar physical geographic positions or/and IP addresses within the set distance threshold value as a user activity address; (2) when a user locates at the same user activity address for a plurality of times in a specific time period (the specific time period is preset by a system, such as the working time period: Monday-Friday, 9: 00-12: 00, 13: 00-17: 00), presuming the meaning of the user activity address to the user (such as a working unit address); (3) the invention obtains the optimal user portrait result based on the user activity address by utilizing the artificial neural network or the inference machine, when the same or similar user activity addresses appear due to different social occupations, different age groups and other factors of the users in the process of constructing the user portrait, the meanings of the user activity addresses to the users may be different (for example, teachers and students use the same address of a school as a common user activity address, but have different meanings), if the data of the ages, the occupations and the like of the users are simultaneously collected to carry out the user portrait, the data collection process is more complicated, the invention considers that the social life habits of the users with different ages and occupations are different (for example, the teachers and the students use the same user activity address in the working period, but can engage in different leisure activities in the leisure time period), and by the training of the artificial neural network or the inference machine, the invention makes full use of the information carried by the physical geographic position or/and the IP address of the user to construct the user portrait, reduces the possible incomplete information and even errors of the user portrait according to the personal information manually filled in when the user registers the system, and effectively improves the accuracy of the user portrait.
The artificial neural network comprises an input layer, an output layer and one or more hidden layers, wherein relevant data of a user activity address collected by the system are processed to obtain user portrait data candidates based on the user activity address, the user portrait data candidates based on the user activity address are used as input data of the artificial neural network and are given to each unit of the input layer, each unit of the hidden layers is weighted summation of each unit of the input layer, the output of the hidden layers is used as input and is transmitted to the output layer, and finally, the optimal user portrait result based on the user activity address is output.
And the inference machine selects a corresponding rule from the knowledge base according to the current content, obtains a corresponding conclusion when the rule is matched with a given fact, stores the conclusion into the comprehensive database, and enables the next rule to be matched if the rule is not matched with the given fact until the matching is successful, so as to reason the conclusion of the problem. Knowledge reasoning methods are divided into forward reasoning, backward reasoning and bidirectional reasoning according to the reasoning direction.
In this embodiment, forward reasoning is employed, i.e., from already obtained user profile data candidates based on user activity addresses, to infer optimal user profile results based on user activity addresses according to matching rules between candidate parameters and user profile tags in the knowledge base.
The embodiment realizes the automatic analysis of the content of the activity which indicates that the user is engaged in when the user physical geographic position or/and the IP address are used in different periods through the steps, thereby determining the user portrait, and the method is a user portrait construction method with higher intelligent degree.
And 5: presetting a time span, monitoring the change of a plurality of similar physical geographic positions or/and IP address data which are repeatedly used by a user for a plurality of times in a fixed time interval and are in the same or in a set distance threshold, and automatically adjusting the user portrait result based on the user activity address according to the data change and the processes of the steps 3 and 4.
The physical geographic position and/or the IP address frequently used by the user in the near future are recorded by setting a dynamic time threshold, and when the recorded data and the data in the previous time interval are greatly changed, the user information is reconstructed, so that the problem that the information change is caused by the user changing work, moving home and the like is solved, and if the information change is not changed in time, wrong user portrait information can appear.
In the embodiment, considering that the user activity address is not constant, when the user activity address changes, often accompanied with the change of the IP address or/and the physical geographic location when the user uses the mobile terminal to surf the internet, for example, when the user changes work, the physical geographic location or/and the IP address also changes during the work period, in order to further improve the user experience and construct an accurate user portrait, the embodiment provides a method for monitoring the change of the physical geographic location or/and the IP address data frequently used by the user recently through a preset time length, reconstructing the user portrait according to the change result, for example, setting the time length to be 2 months, when the user does not use the original physical geographic location or/and the IP address continuously for 2 months (or does not use the physical geographic location and/or the IP address mostly during 2 months), and when the user activity address uses a new physical geographic location or/and/or IP address during the work period, and automatically taking the new physical geographic position or/and IP address used in the working period as the working address of the user.
Further, a method of user portrayal based on user activity address, the slot tag comprising: work hours, dining hours, sleeping hours, leisure hours, weekend hours, holidays, roaming to a foreign place; the user profile tag includes: companies, institutions, schools, home addresses, dormitories, hotels; the time interval labels are preset by a system and can be adjusted by a user according to the actual conditions of the user, and each time interval label sets the corresponding time interval length.
For example, the storage unit corresponding to the server is pre-provided with a preset time interval tag provided by a service provider, and the preset working time interval is as follows: monday to friday, 9: 00-12: 00, 13: 00-17: 00, and the working time of a specific user varies due to different work, thereby allowing the user to customize the use.
The present embodiment, based on the creation of a user portrait tag library for a period tag and a user activity address of a frequency threshold, may use the following principles: the user continuously accumulates for 50 times within 30 times or 2 months in the working time period of a working day, and sets user portrait labels as companies, units and schools at a plurality of similar physical geographic positions or/and IP addresses which are the same or within a set distance threshold; the user continuously accumulates for 60 times within 90 or 3 months from 6 hours in the morning to 6 hours in the morning, and sets user portrait labels as home addresses and dormitories at a plurality of similar physical geographic positions or/and IP addresses within a set distance threshold; the user sets the user portrait label as a home address, a hotel, a tourist spot and the like at the dining time of a major holiday and at a plurality of similar physical geographic positions or/and IP addresses within a set distance threshold.
Because the social activities that the user can engage in are more, the invention adopts each time interval label to set the corresponding time interval length according to the experience, for example, the working time interval is divided into (Monday-Friday, 9: 00-12: 00, 13: 00-17: 00), the bedtime interval is divided into (23: 00-6: 00 every night), and the like, the physical geographic location information of the working time interval corresponds to the working address, the physical geographic location information of the bedtime interval corresponds to the home address or the temporary address, the IP address of the dining time interval and/or the precise place name corresponding to the physical geographic location information correspond to favorite or frequently-going restaurants, and the like, and the corresponding correlation relations are used for constructing the user portrait and are not precise enough, therefore, the frequency of the repeated occurrence of the information in the corresponding time interval needs to be calculated, and when the repeated frequency is greater than the set frequency threshold, the corresponding user portrait label is matched.
Further, a method for user representation based on user activity address, the method for the close physical geographic location or/and IP address distance threshold in step 2 can be preset by the system, and can also be manually set by the user.
The embodiment allows the user to set the distance threshold according to the actual situation of the user, so as to solve the problem that the range of the activity place is different when the user is engaged in the actual social activity.
Further, a method for user representation based on user activity address, the method for obtaining physical geographic location or/and IP address when user accesses internet by using mobile terminal, includes:
judging the internet access mode of a user mobile terminal, if a user uses a mobile communication network to access the internet, acquiring the communication base station information accessed by the mobile phone of the user, and acquiring the physical geographic position corresponding to the communication base station;
if the user uses WiFi to surf the internet, acquiring an IP address corresponding to the WiFi used by the user, and acquiring a physical geographic position corresponding to the IP address;
and if the user allows to use the GPS positioning, directly acquiring the corresponding physical geographic position according to the GPS positioning.
After obtaining the physical geographic location or/and the IP address used by the user, querying a corresponding precise place name, including: and obtaining an accurate place corresponding to the physical geographic position or/and the IP address through network search according to the obtained physical geographic position or/and the IP address, and obtaining the accurate place name.
When a user uses the mobile terminal to surf the internet, the mobile terminal can be connected to a public network in various modes, the physical geographic position information corresponding to a communication base station accessed by the user can be obtained through analysis by using the mobile communication network, the corresponding physical geographic position information can also be obtained according to an IP address by using WiFi, and if the user allows a system to collect GPS positioning information of the mobile terminal when using the mobile terminal, more accurate physical geographic position information can be obtained.
Specifically, in this embodiment, for example, the system is allowed to record the physical geographic location or/and the IP address of the user when using the mobile terminal in about 2 months, the system uploads the collected related information to the server and stores the information through the physical geographic location or/and IP address collection device corresponding to the mobile terminal, when the system collects enough data, the user representation can be automatically constructed, and first, the collected physical geographic location or/and IP address used by the user in about 2 months is classified according to time periods, such as: physical geographic locations or/and IP addresses used during work hours, physical geographic locations or/and IP addresses used during bedtime hours, etc., then calculating the frequency of repeated use of a physical geographic position or/and IP address for a plurality of times in each time period, if user a had accumulated 55 times during the work period in the last 2 months using physical geographical location information labeled "ne 398 of XX city), according to the user portrait label library of user activity addresses based on the time interval labels and the frequency threshold, the user portrait label corresponding to the physical geographic position information used in the working period is the 'working address', the frequency threshold for judging the corresponding matching relationship is established is 40 times, because the actual use times are greater than the preset frequency threshold, a corresponding matching relationship and a candidate user image are established, that is, the following dimension information is added to the candidate user image: the working address is as follows: XX east wind road 398, further, if it is found through network search that the accurate place information corresponding to "XX east wind road 398" is "XX second people hospital", establishing a matching relationship between the accurate place name and a work unit label, and constructing a candidate user image, that is, adding the following dimension information in the candidate user image: the working unit is as follows: finally, if there are other candidate items in the second people hospital in XX, the optimal user portrait result is obtained through artificial neural network training (in this embodiment, it is mainly explained how to obtain the precise place name, so that it is not specifically explained whether it is a precise place or several precise places, and when there are multiple precise places, the processing method is the same as that in the foregoing embodiment).
Referring to FIG. 2, a user representation apparatus based on a user activity address comprises:
the device comprises a physical geographic position or/and IP address acquisition device, a server and a server, wherein the physical geographic position or/and IP address acquisition device is used for acquiring the physical geographic position or/and IP address when a user uses the mobile terminal to surf the internet, and corresponding information of use time and use time length and uploading the information to the server;
the device comprises a distance threshold setting device for setting and judging a plurality of close physical geographic positions or/and IP addresses to be a distance threshold corresponding to a user activity address according to the actual situation of a user;
and the user portrait result receiving device is used for receiving the user portrait result which is obtained by the server after calculation and is based on the user activity address, and providing further personalized service for the user according to the user portrait result based on the user activity address.
The specific implementation mode of the device can be a mobile terminal (a mobile phone, a tablet computer, a vehicle-mounted terminal and the like) provided with an APP (application program) comprising the functions of the above parts, and after the mobile terminal is connected with a server, the user portrait function of the user of the mobile terminal can be completed by the above parts.
Referring to FIG. 2, a system for user representation based on user activity address, comprising:
a mobile terminal;
a server, comprising:
a user physical geographic position or/and IP address storage unit, which is used for storing the physical geographic position or/and IP address when the user registered on the server uses the corresponding mobile terminal, and the time information when the user uses the IP address or/and is at the physical geographic position;
the user portrait label library unit is used for storing time periods and frequency of user activity addresses corresponding to a plurality of similar physical geographic positions or/and IP addresses of a user in the same or within a set distance threshold value, and judging the corresponding relation between the time period labels, the frequency threshold value and the user activity addresses and the user portrait labels;
the user portrait construction unit is used for obtaining a user portrait based on a user activity address by inference according to a user physical geographic position or/and an IP address, and the structure of the user portrait construction unit is shown in FIG. 3 and comprises: the system comprises a module for calling a physical geographic position or/and an IP address of a user, a time period analysis device for analyzing use time and use time length information corresponding to the physical geographic position or/and the IP address when the user uses a mobile terminal to surf the internet, a frequency calculation device for calculating the user activity address repeatedly used by the user for multiple times in a fixed time period, and an artificial neural network training model or a user portrait inference device consisting of a knowledge base and an inference machine;
and the user portrait output device is used for outputting the user portrait based on the user activity address obtained by the user portrait construction unit.
The system calls the information of the physical geographic position or/and the IP address used by the user through the server, calculates and obtains the user portrait according to the data of the physical geographic position or/and the IP address used by the user in different periods, and has higher intelligent degree.
Further, the system for user representation based on user activity address comprises:
the user address based user representation apparatus further comprises: the time length presetting device is used for customizing a user data monitoring time length by a user so as to monitor the change of the physical geographic position or/and the IP address.
The server further comprises:
the user portrait adjusting unit is used for monitoring the data change condition of the user using the physical geographic position or/and the IP address in a fixed time period within a preset time length, and if the data change exceeds a preset threshold value, adjusting the user portrait based on the user activity address;
the mobile terminal network judging device is used for analyzing the internet access mode of the mobile terminal connected to the server;
the physical geographic position information determining device is used for acquiring a corresponding physical geographic position or/and an IP address according to the internet access mode of the mobile terminal and finally acquiring physical geographic position information;
and the accurate place network searching device is used for determining the accurate place name corresponding to the physical geographic position or/and the IP address through network searching.
The parts not involved in the present invention are the same as or can be implemented using the prior art.
The above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, and any person skilled in the art can modify or easily think the technical solutions described in the above-mentioned embodiments within the technical scope of the present invention, or make equivalent substitutions on some technical features, and these modifications, changes or substitutions should not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and shall be covered within the protection scope of the present invention.
Claims (5)
1. A method for user portrayal based on a user activity address, comprising:
step 1:
(1.1) presetting a user portrait label library of user activity addresses based on a time interval label and a frequency threshold, wherein the user activity addresses refer to a plurality of similar physical geographic positions or/and IP addresses in the same or within a set distance threshold;
(1.2) establishing an artificial neural network in advance through candidate user image data samples based on user activity addresses, and training the neural network by using a learning training module until the network converges; or establishing a knowledge base through a candidate user portrait data sample based on a user activity address, wherein the knowledge base is a knowledge map;
step 2: presetting the close physical geographic position or/and the IP address distance threshold;
and step 3:
(3.1) calculating time interval labels and frequency data of a plurality of similar physical geographic positions or/and IP addresses of the user in the same or within a set distance threshold;
the specific implementation process of the step (3.1) is as follows: acquiring a physical geographic position or/and an IP address when a user surfs the internet by using a mobile terminal, acquiring a corresponding time interval label according to a time interval when the user uses the same or a plurality of similar physical geographic positions or/and IP addresses within a set distance threshold, and counting the frequency of using the corresponding physical geographic position or/and IP address by the user within the time interval;
(3.2) matching a preset user portrait label library of the user activity address based on the time interval label and the frequency threshold, and calculating a user portrait data candidate based on the user activity address; if 1 candidate is obtained, the candidate is adopted as a user portrait result; if multiple candidates are obtained, entering step 4;
the specific implementation process of the step (3.2) is as follows: matching the user portrait label library of the user activity address based on the time period label and the frequency threshold value preset in the step (1.1) according to the time period label and the frequency data obtained in the step (3.1), taking the corresponding label in the user portrait label library as a user portrait label candidate based on the user activity address, and when the user uses different physical geographic positions or/and IP addresses in the same time period, obtaining a plurality of user portrait label candidates based on the user activity address, and combining the user portrait label candidate based on the user activity address, the corresponding user activity address, the time period label of the user activity address used by the user, the time length of one-time use, the use frequency, the number of times and the accumulated time length in the time period into a user portrait data candidate based on the user activity address;
and 4, step 4: inputting the user portrait data candidate item based on the user activity address obtained in the step (3) into the artificial neural network trained in the step (1.2) to obtain an optimal user portrait result based on the user activity address; or matching corresponding rules from the knowledge base through an inference engine until matching is successful, and obtaining an optimal user portrait result based on a user activity address;
the specific implementation process of the step 4 is as follows: inputting the user portrait data candidate item processed in the step (3) based on the user activity address into an artificial neural network, wherein an input layer of the artificial neural network recognizes a label candidate item of the user portrait based on the user activity address, and a corresponding user activity address, a time period label of the user activity address used by the user, a time length of one-time use, a frequency of use in a time period, a frequency of use in the time period, and an accumulated time length as input parameters, and the input parameters are transmitted to an output layer through a hidden layer, so that an optimal user portrait result based on the user activity address is obtained by the output layer; or matching corresponding rules from the knowledge base through an inference engine until matching is successful, and obtaining an optimal user portrait result based on a user activity address;
and 5: presetting a time span, monitoring the change of a plurality of similar physical geographic positions or/and IP address data which are repeatedly used by a user for a plurality of times in a fixed time interval and are in the same or in a set distance threshold, and automatically adjusting the user portrait result based on the user activity address according to the data change and the processes of the steps 3 and 4.
2. The method of claim 1, wherein the time period label comprises: work hours, dining hours, sleeping hours, leisure hours, weekend hours, holidays, roaming to a foreign place; the user profile tag includes: companies, institutions, schools, home addresses, dormitories, hotels; the time interval labels are preset by a system and can be adjusted by a user according to the actual conditions of the user, and each time interval label sets the corresponding time interval length.
3. The method according to claim 1, wherein the method of the proximity physical geographic location or/and the IP address distance threshold in step 2 can be preset by a system or manually set by a user.
4. The method according to claim 3, wherein the method for obtaining the physical geographic location or/and the IP address of the user when the user accesses the internet by using the mobile terminal comprises:
judging the internet access mode of a user mobile terminal, if a user uses a mobile communication network to access the internet, acquiring the communication base station information accessed by the mobile phone of the user, and acquiring the physical geographic position corresponding to the communication base station;
if the user uses WiFi to surf the internet, acquiring an IP address corresponding to the WiFi used by the user, and acquiring a physical geographic position corresponding to the IP address;
and if the user allows to use the GPS positioning, directly acquiring the corresponding physical geographic position according to the GPS positioning.
5. The method according to claim 4, wherein after obtaining the physical geographic location or/and the IP address used by the user, querying the corresponding precise location name comprises: and obtaining an accurate place corresponding to the physical geographic position or/and the IP address through network search according to the obtained physical geographic position or/and the IP address, and obtaining the accurate place name.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710930889.7A CN107704586B (en) | 2017-10-09 | 2017-10-09 | User portrait method, device and system based on user activity address |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710930889.7A CN107704586B (en) | 2017-10-09 | 2017-10-09 | User portrait method, device and system based on user activity address |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107704586A CN107704586A (en) | 2018-02-16 |
| CN107704586B true CN107704586B (en) | 2020-02-07 |
Family
ID=61184811
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710930889.7A Active CN107704586B (en) | 2017-10-09 | 2017-10-09 | User portrait method, device and system based on user activity address |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107704586B (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110300084B (en) * | 2018-03-22 | 2023-09-01 | 北京京东尚科信息技术有限公司 | IP address-based portrait method and apparatus, electronic device, and readable medium |
| CN108804507B (en) * | 2018-04-16 | 2019-07-30 | 北京嘀嘀无限科技发展有限公司 | The address location determining method and system of user |
| CN110457359B (en) * | 2018-05-04 | 2024-03-08 | 拉萨经济技术开发区凯航科技开发有限公司 | Correlation analysis method |
| CN108615199A (en) * | 2018-05-11 | 2018-10-02 | 国家计算机网络与信息安全管理中心 | The User Activity track method for digging of forum's registration scenarios is disclosed based on internet |
| CN110737690A (en) * | 2018-07-03 | 2020-01-31 | 百度在线网络技术（北京）有限公司 | User label mining method and device, computer equipment and storage medium |
| CN109195219B (en) * | 2018-09-17 | 2021-01-26 | 每日互动股份有限公司 | Method for determining position of mobile terminal by server |
| CN109660980B (en) * | 2018-09-27 | 2023-05-12 | 深圳壹账通智能科技有限公司 | Method, device, equipment and medium for constructing user portrait based on history connection WiFi |
| CN111127064B (en) * | 2018-11-01 | 2023-08-25 | 百度在线网络技术（北京）有限公司 | Method and device for determining social attribute of user and electronic equipment |
| CN109769210A (en) * | 2018-11-23 | 2019-05-17 | 亚信科技（中国）有限公司 | User Activity Regional Similarity judgment method, device, computer equipment |
| CN109783597B (en) * | 2019-01-29 | 2020-12-29 | 江苏满运软件科技有限公司 | Method, device, electronic equipment and storage medium for determining fixed position of driver |
| CN110324327B (en) * | 2019-06-20 | 2021-07-13 | 国家计算机网络与信息安全管理中心 | User and server IP address calibration device and method based on specific enterprise domain name data |
| CN110365755A (en) * | 2019-06-28 | 2019-10-22 | 深圳数位传媒科技有限公司 | A kind of information recommendation method and device triggered in real time based on key scenes |
| CN110930180B (en) * | 2019-11-04 | 2023-07-28 | 苏宁云计算有限公司 | Data analysis method, system and computer equipment based on regional member marketing scene |
| CN110852372B (en) * | 2019-11-07 | 2022-05-31 | 北京爱笔科技有限公司 | Data association method, device and equipment and readable storage medium |
| CN111639253B (en) * | 2020-05-22 | 2023-08-01 | 北京百度网讯科技有限公司 | Data weight judging method, device, equipment and storage medium |
| CN111815375B (en) * | 2020-09-04 | 2021-01-08 | 北京悠易网际科技发展有限公司 | User portrayal method and device in advertisement putting |
| CN113449114A (en) * | 2020-12-31 | 2021-09-28 | 中国科学技术大学智慧城市研究院(芜湖) | Method for constructing natural human life cycle holographic image based on knowledge graph |
| CN114071366B (en) * | 2022-01-17 | 2022-05-24 | 北京融信数联科技有限公司 | Figure portrait depicting method and system in combination with knowledge graph and readable storage medium |
| CN115190110B (en) * | 2022-07-18 | 2024-01-09 | 北京抖音信息服务有限公司 | Geographic position determining method and device |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239327B (en) * | 2013-06-17 | 2017-11-07 | 中国科学院深圳先进技术研究院 | A kind of mobile Internet user behavior analysis method and device based on positional information |
| CN104933126A (en) * | 2015-06-11 | 2015-09-23 | 百度在线网络技术（北京）有限公司 | Activity recommending method based on electronic map, activity recommending device based on electronic map and activity recommending system based on electronic map |
| CN106548299A (en) * | 2016-11-24 | 2017-03-29 | 杭州领点科技有限公司 | A kind of data analysis system and method based on user profile |
| CN107016042B (en) * | 2017-02-13 | 2023-07-04 | 南京安讯科技有限责任公司 | Address information verification system based on user position log |
-
2017
- 2017-10-09 CN CN201710930889.7A patent/CN107704586B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107704586A (en) | 2018-02-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107704586B (en) | User portrait method, device and system based on user activity address | |
| Ahas et al. | Everyday space–time geographies: using mobile phone-based sensor data to monitor urban activity in Harbin, Paris, and Tallinn | |
| US20220148010A1 (en) | Generating an activity inference model from contextual data | |
| Zhao et al. | Understanding the bias of call detail records in human mobility research | |
| CN106021363B (en) | course recommendation method and device | |
| Montoliu et al. | Discovering places of interest in everyday life from smartphone data | |
| KR102121361B1 (en) | Method and device for identifying the type of geographic location where the user is located | |
| CN107851243B (en) | Inferring physical meeting location | |
| Ferrari et al. | Discovering daily routines from google latitude with topic models | |
| CN104737523A (en) | Managing a context model in a mobile device by assigning context labels for data clusters | |
| CN107613084B (en) | Method, device and system for automatically grouping contacts in address list | |
| WO2013091302A1 (en) | Intelligent search method and device | |
| CN111191133B (en) | Service search processing method, device and equipment | |
| Doyle et al. | Population mobility dynamics estimated from mobile telephony data | |
| JP2010009315A (en) | Recommended store presentation system | |
| Adams et al. | Sensing and using social context | |
| CN104021466B (en) | Multi-user's schedule method and apparatus | |
| CN106850106A (en) | Radio environment map datum collection system and method based on mobile intelligent perception | |
| CN113454669A (en) | Characterizing a place by user visited features | |
| Heikinheimo et al. | Detecting country of residence from social media data: a comparison of methods | |
| KR20200133976A (en) | Contents Curation Method and Apparatus thereof | |
| Whitney et al. | Developments in big data for park management: a review of mobile phone location data for visitor use management | |
| Papliatseyeu et al. | Mobile habits: Inferring and predicting user activities with a location-aware smartphone | |
| CN112765374A (en) | Education resource screening system and method for information push | |
| JP2011154484A (en) | Method, device and program for deciding behavioral situation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |