CN114461926A - Precise matching method for public accumulation fund payment person house purchasing based on double collaborative filtering - Google Patents
Precise matching method for public accumulation fund payment person house purchasing based on double collaborative filtering Download PDFInfo
- Publication number
- CN114461926A CN114461926A CN202210129340.9A CN202210129340A CN114461926A CN 114461926 A CN114461926 A CN 114461926A CN 202210129340 A CN202210129340 A CN 202210129340A CN 114461926 A CN114461926 A CN 114461926A
- Authority
- CN
- China
- Prior art keywords
- commodity
- historical
- house
- payer
- rec
- 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
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
- G06F16/90—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
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a precise matching method for buying rooms of a public accumulation fund payment person based on double collaborative filtering. The method comprises the following steps: based on historical payer data and commodity room data, firstly, using cooperative filtering based on a payer to obtain a commodity room list Rec _ user matched according to the house purchasing demand of the to-be-detected payer; then, using collaborative filtering based on commodity houses to obtain a commodity house list Rec _ house matched with the to-be-detected payer; and finally, removing the weight of the commodity room list Rec _ user and Rec _ house, and taking the intersection of the commodity room list Rec _ user and the commodity room list ghouse which can use the public accumulation fund loan to purchase the houses as a commodity room list Rec _ list which is finally pushed to the to-be-detected payer. The invention can recommend accurate house information to the user who purchases house using the public accumulation loan quickly and accurately.
Description
Technical Field
The invention relates to the field of accurate matching, in particular to a precise matching method for buying a house by a public accumulation fund payment person based on double collaborative filtering.
Background
In recent years, the informatization management level of house public accumulation fund management centers (hereinafter referred to as "Chongqing public accumulation fund centers") in Chongqing cities is continuously improved, the business volume is gradually increased, and the personal information of public accumulation fund payers interacts with the information of different service channels to generate large data with multiple sources, isomerism and large scale. How to scientifically mine and utilize mass data resources of a Chongqing public accumulation fund center and provide a higher-quality service for public accumulation fund payers; how to help the depositor to conveniently, quickly and accurately obtain the most needed information from thousands of data sources and massive information becomes a new problem of highly paying attention and deeply exploring in the Chongqing public accumulation fund center.
With the rapid development of the internet industry, new generation information technologies such as big data, cloud computing and artificial intelligence are widely applied to different industries and fields, and the information amount is increased explosively. Under the environment of information overload, how to accurately match information interested by a user with the user becomes a key technical problem in the field of information retrieval. The personalized recommendation system is a key technology for solving overload of information quantity, and is successfully applied to industries and fields of e-commerce, social contact, short videos, news and the like. The technology aims to analyze the interests of user groups by applying technologies such as big data, artificial intelligence and the like and combining with mass data resources related to the industry, connect users with interested information, help the users to quickly and accurately acquire the interested information and realize personalized accurate recommendation.
The Chongqing public accumulation fund center conforms to the development of the era, and in order to further improve the precise service management level, a house purchasing precise matching method is constructed by utilizing the personalized recommendation system technology, so that the problem of solving and improving houses by vast payers can be better helped, and more precise, more efficient, more intelligent and more attentive public accumulation fund service experience can be obtained.
Currently, there is no recommendation method for providing market house purchase information to a payer who can purchase a house by using a public accumulation fund loan, so a method is urgently needed to solve the problem of accurate matching of house purchases.
Disclosure of Invention
Aiming at the problems in the prior art, the technical problems to be solved by the invention are as follows: a precise house matching method based on the purchase of houses by the public accumulation fund is constructed.
In order to solve the technical problems, the invention adopts the following technical scheme:
a precise matching method for a public deposit payer house purchase based on double collaborative filtering is characterized by comprising the following steps: the method comprises the following steps:
s100: the method comprises two collaborative filtering models, wherein one collaborative filtering model is based on a payer, and the other collaborative filtering model is based on a commodity house;
s200: constructing a collaborative filtering model based on a payer, and comprising the following steps;
s210: acquiring public deposit history payer information user and commodity house information house purchased by a public deposit history payer;
s220: defining a historical payer information matrix Vuser by using the historical payer information of the public deposit, defining a historical commodity room information matrix Vhouse by using the commodity room information purchased by the historical payer of the public deposit, wherein the definition expressions are respectively as follows:
historical payer information matrixWherein n represents the total number of the historical payers, xn ... znThe method comprises the steps of representing relevant information characteristics of n historical payers;
historical commodity room information matrixWherein m represents the total number of the historical commodity rooms, am... bmThe related information characteristics of the m historical commodity rooms are represented;
s230: selecting information of a payer to be detected, and calculating similarity scores, user, between the information of the payer to be detected and each piece of historical payer informationiShowing the information of the ith to-be-detected payer,representing the information of the h-th historical payer in the n historical payers;
s240: the similarity scores are arranged in descending order to obtain historical paymentPeople similarity score ranking UserScorenAnd historical Commodity Room List Rec _ user, UserScorenThe similarity scores in the Rec _ user are in one-to-one correspondence with the commodity rooms in the Rec _ user, and the similarity scores of the historical payers are ranked to UserScorenAs shown in formula (1), the corresponding historical commodity room list Rec _ user is shown in formula (2):
UserScoren=Sim(topu1,topu2,...,topun) (1)
where u represents the descending ranking, topunIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scoreun) Denotes topunA corresponding similarity score;
Rec_user=V(topu1,topu2,...,topun) (2)
wherein, V (top)un) Representative sequence number topunA corresponding commodity room;
s300: constructing a collaborative filtering model based on a commodity house, and comprising the following steps;
s310: ranking UserScore according to similarity scores of historical payersnThe public deposit historical payer corresponding to the highest score in the group is used as the historical payer most similar to the to-be-detected payer, and the commodity house purchased by the most similar historical payer is used as the initial push commodity house;
s320: calculating similarity scores between the initial pushed commodity room and the rest m-1 historical commodity rooms;
s330: the similarity scores are arranged according to the scores in a descending order to obtain the commodity house similarity score ranking HouseScorenAnd a list of commercial houses Rec _ house, House corenThe similarity scores in the Rec _ house are in one-to-one correspondence with the commodity rooms in the Rec _ house, and the similarity scores are ranked as HouseStorenAs shown in formula (3), the commodity room list Rec _ house is shown in formula (4):
HouseScorem=Sim(toph1,toph2,...,tophm) (3)
wherein, tophmIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scorehm) Represents tophmA corresponding similarity score;
Rec_house=V(toph1,toph2,...,tophm) (4)
wherein, V (top)hm) Representative sequence number tophmA corresponding commodity room;
s400: acquiring commodity house information ghouse which can be purchased by using a public accumulation fund loan;
s500: and carrying out deduplication processing on Rec _ user and Rec _ house to obtain a commodity room list Rec _ temp after deduplication, wherein the expression is as follows:
Rec_temp=Rec_user ∪ Rec_house (5);
s600: intersecting the Rec _ temp and the ghouse to obtain a commodity room list Rec _ list which is finally pushed to a to-be-detected payer, wherein the specific calculation expression is as follows:
Rec_list=ghouse ∩ Rec_temp。
preferably, the specific method for calculating the similarity score between the to-be-detected payer information and each historical payer information in S230 is a Jaccard (Jaccard) similarity coefficient method, and the specific expression is as follows:
wherein the content of the first and second substances,the information matrix of the h-th historical payer in the n historical payers is represented, wherein h is 1,2, … n, VuseriAnd (3) an information matrix representing the ith to-be-detected payer.
Preferably, the similarity calculation in S320 uses a cosine similarity calculation method, and the specific expression is as follows:
wherein the content of the first and second substances,representing an initial pushed merchandising room information matrix,and the information matrix represents the h-th historical commodity room in the m historical commodity rooms.
Compared with the prior art, the invention has at least the following advantages:
1. the invention can recommend the building capable of using the public accumulation fund loan for the public accumulation fund depositor, and the model accuracy rate reaches 84.1 percent.
2. The invention can provide better service for the housing demand of the public deposit depositor and improve the housing purchasing efficiency.
3. The invention explores a solution for accurately matching requirements of the house purchasing of the payer in the accumulation fund industry for the first time, and has exploration demonstration function in the accumulation fund industry.
Drawings
Fig. 1 is an overall framework of the present invention.
Detailed Description
The present invention is described in further detail below.
The invention describes a precise matching method for buying rooms of a public accumulation fund payment person based on double collaborative filtering. The core idea of the invention is that a commodity room list to be matched generated by a model is automatically predicted by a given payer which accords with a room purchasing policy and has room purchasing ability and a trained accurate matching model, and the list is pushed to the payer.
Based on historical payer data and commodity room data, the method firstly uses cooperative filtering based on a payer to obtain a commodity room list Rec _ user matched according to the house purchasing demand of a to-be-detected payer; then, using collaborative filtering based on commodity houses to obtain a commodity house list Rec _ house matched with the to-be-detected payer; and finally, removing the weight of the commodity room list Rec _ user and Rec _ house, and taking the intersection of the commodity room list Rec _ user and the commodity room list ghouse which can use the public accumulation fund loan to purchase the houses as a commodity room list Rec _ list which is finally pushed to the to-be-detected payer.
The invention comprises two components. Component oneThrough calculating the user of the to-be-detected payment personiPaying the deposit with historyIs given by the similarity score ofObtaining a user score for the similarity score ranking of the payersnFinally mapping to obtain a corresponding commodity room list Rec _ user; commodity house with two-way module through calculationWith historical other commodity roomsDegree of similarity ofHouseStore for obtaining commodity house similarity score rankingnAnd then generating a commodity house list Rec _ house matched with the to-be-detected payer. And finally, removing the duplication of the commodity house list Rec _ user and Rec _ house, and taking the intersection of the commodity house list Rec _ user and the geocouse as the commodity house list Rec _ list finally pushed to the to-be-detected depositor.
Referring to fig. 1, a precise matching method for buying a house by a public deposit payer based on double collaborative filtering is characterized in that: the method comprises the following steps:
s100: the method comprises two collaborative filtering models, wherein one collaborative filtering model is based on a payer, and the other collaborative filtering model is based on a commodity house;
s200: constructing a collaborative filtering model based on a payer, and comprising the following steps;
s210: acquiring public deposit history payer information user and commodity house information house purchased by a public deposit history payer;
s220: defining a historical payer information matrix Vuser by using the historical payer information of the public deposit, defining a historical commodity room information matrix Vhouse by using the commodity room information purchased by the historical payer of the public deposit, wherein the definition expressions are respectively as follows:
historical payer information matrixWherein n represents the total number of the historical payers, xn ... znThe method comprises the steps of representing relevant information characteristics of n historical payers;
historical commodity room information matrixWherein m represents the total number of the historical commodity rooms, am... bmThe related information characteristics of the m historical commodity rooms are represented;
s230: selecting information of a payer to be detected, and calculating similarity scores, user, between the information of the payer to be detected and each piece of historical payer informationiThe information of the ith payer to be detected is shown,representing the information of the h-th historical payer in the n historical payers;
the specific method for calculating the similarity score between the to-be-detected payer information and each historical payer information in S230 is a Jaccard (Jaccard) similarity coefficient method, and the specific expression is as follows:
wherein the content of the first and second substances,the information matrix of the h-th historical payer in the n historical payers is represented, wherein h is 1,2, … n, VuseriAnd expressing the information matrix of the ith to-be-detected payer.
S240: the similarity scores are sorted in a descending order to obtain historical payer similarity score ranking UserScorenAnd historical Commodity Room List Rec _ user, UserScorenThe similarity scores in the Rec _ user are in one-to-one correspondence with the commodity rooms in the Rec _ user, and the similarity scores of the historical payers are ranked to UserScorenAs shown in formula (1), the corresponding historical commodity room list Rec _ user is shown in formula (2):
UserScoren=Sim(topu1,topu2,...,topun) (1)
where u represents the descending ranking, topunIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scoreun) Denotes topunA corresponding similarity score;
Rec_user=V(topu1,topu2,...,topun) (2)
wherein, V (top)un) Representative sequence number topunA corresponding commodity room;
s300: constructing a collaborative filtering model based on a commodity house, and comprising the following steps;
s310: ranking UserScore according to similarity scores of historical payersnThe public deposit historical payer corresponding to the highest score in the group is used as the historical payer most similar to the to-be-detected payer, and the commodity house purchased by the most similar historical payer is used as the initial push commodity house;
s320: calculating similarity scores between the initial pushed commodity room and the rest m-1 historical commodity rooms;
the similarity calculation in S320 uses a cosine similarity calculation method, and the specific expression is as follows:
wherein the content of the first and second substances,representing an initial pushed merchandising room information matrix,and the information matrix represents the h-th historical commodity room in the m historical commodity rooms.
S330: the similarity scores are arranged in a descending order according to the scores to obtain a commodity room similarity score ranking HouseScorenAnd a list of commercial houses Rec _ house, House corenThe similarity scores in the Rec _ house are in one-to-one correspondence with the commodity rooms in the Rec _ house, and the similarity scores are ranked as HouseStorenAs shown in formula (3), the commercial room list Rec _ house is shown in formula (4):
HouseScorem=Sim(toph1,toph2,...,tophm) (3)
wherein, tophmIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scorehm) Denotes tophmA corresponding similarity score;
Rec_house=V(toph1,toph2,...,tophm) (4)
wherein, V (top)hm) Representative sequence number tophmA corresponding commodity room;
s400: acquiring commodity house information ghouse which can be purchased by using a public accumulation fund loan;
the commodity house information gHouse is used for defining a commodity house information matrix VgHouse which can be purchased by a public fund loan, and the expression is defined as follows: commodity housing information matrixWherein j represents the total number of the commodity rooms, pj ... qjThe method comprises the steps of representing relevant information characteristics of j commodity rooms, wherein all relevant information about the commodity rooms is contained in a Vghouse information matrix;
s500: and carrying out deduplication processing on Rec _ user and Rec _ house to obtain a commodity room list Rec _ temp after deduplication, wherein the expression is as follows:
Rec_temp=Rec_user ∪ Rec_house (5);
s600: intersecting the Rec _ temp and the ghouse to obtain a commodity room list Rec _ list which is finally pushed to a to-be-detected payer, wherein the specific calculation expression is as follows:
Rec_list=ghouse ∩ Rec_temp。
rec _ list also has a good house information matrix, which contains all the relevant house information about the good houses in the recommendation list.
Experimental verification
The method utilizes a collaborative filtering algorithm and designs three models for experimental comparison and analysis according to different collaborative filtering mechanisms. Model one: the basic idea is that the most similar payer is found in the history database with the personal information, and the commodity house information is pushed to the payer meeting the condition of the public deposit loan; model two: the basic idea is that the commodity house most similar to the commodity house is found out by finding out the commodity house information purchased by the payer who uses the public deposit loan, and the similar commodity house is pushed to the payer; and (3) model III: and combining the historical similar payer information and the similar commodity room information, constructing an accurate matching model of the commodity room, and pushing the commodity room for the payer.
The data used in the experiment are from a database of a Chongqing public accumulation fund center, and after data preprocessing, the final data set comprises 109816 pieces of payer data and 3329 pieces of building data which can use a public accumulation fund loan.
The experiment adopts an automatic evaluation method, selects the Accuracy @ k index in a recommendation system as an evaluation index, wherein,k represents the recommended number, Num _ acc (k) represents the correct number of the first k results.
The effect pair ratios of the three synergistic filtration models selected in the experiment are shown in table 1. Under the condition of pushing a top-10 commodity room, the matching accuracy of the first model is 0.792, the matching accuracy of the second model is 0.644, and the matching accuracy of the third model is 0.841; it can be seen that the model three has the best effect compared with other models, so the model three is used as the model finally selected by the method.
TABLE 1 comparison of Effect of three synergistic filtration models
Method | Accuracy@k |
Model one | 0.792 |
Model two | 0.644 |
Model III | 0.841 |
In addition, the method subdivides three classes according to the difference of the algorithms used for similarity calculation. The algorithm used in the experiment uses cosine similarity to calculate similarity when searching for similar houses, and a similarity calculation method II comprises the following steps: calculating similarity using euclidean distance when finding similar houses; similarity algorithm three: the pearson correlation coefficient is used to calculate similarity when finding similar houses.
The final comparison results are shown in table 2, where the accuracy of the similarity algorithm used in the present invention is 0.841, the accuracy of the similarity algorithm two is 0.806, and the accuracy of the similarity algorithm three is 0.823. It can be seen that the similarity algorithm used in the present invention is optimal.
TABLE 2 comparison of the effects of the three similarity algorithms
Method | Accuracy@k |
The invention | 0.841 |
Similarity calculation method two | 0.806 |
Similarity calculation method three | 0.823 |
In short, experimental results prove that the accurate house purchasing matching method for the house public accumulation fund payers based on the double-collaborative filtering, disclosed by the invention, can effectively push commodity houses meeting house purchasing requirements for the public accumulation fund payers. Meanwhile, the method can be applied to actual working scenes, and contributes to intelligent development of the Chongqing public accumulation fund center.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.
Claims (3)
1. A precise matching method for a public deposit payer house purchase based on double collaborative filtering is characterized by comprising the following steps: the method comprises the following steps:
s100: the method comprises two collaborative filtering models, wherein one collaborative filtering model is based on a payer, and the other collaborative filtering model is based on a commodity house;
s200: constructing a collaborative filtering model based on a payer, and comprising the following steps;
s210: acquiring public deposit history payer information user and commodity house information house purchased by a public deposit history payer;
s220: defining a historical payer information matrix Vuser by using the historical payer information of the public deposit, defining a historical commodity room information matrix Vhouse by using the commodity room information purchased by the historical payer of the public deposit, wherein the definition expressions are respectively as follows:
historical payment creditInformation matrixWherein n represents the total number of the historical payers, xn…znThe method comprises the steps of representing relevant information characteristics of n historical payers;
historical commodity room information matrixWherein m represents the total number of the historical commodity rooms, am…bmThe related information characteristics of the m historical commodity rooms are represented;
s230: selecting the information of the payers to be detected, and calculating the similarity score, user, between the information of the payers to be detected and the information of each historical payersiShowing the information of the ith to-be-detected payer,representing the information of the h-th historical payer in the n historical payers;
s240: the similarity scores are sorted in a descending order to obtain historical payer similarity score ranking UserScorenAnd historical Commodity Room List Rec _ user, UserScorenThe similarity scores in the Rex _ user are in one-to-one correspondence with commodity houses in the Rex _ user, and the similarity scores of the historical payers are ranked to UserScorenAs shown in formula (1), the corresponding historical commodity room list Rec _ user is shown in formula (2):
UserScoren=Sim(topu1,topu2,...,topun) (1)
where u denotes the descending ranking, topunIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scoreun) Denotes topunA corresponding similarity score;
Rec_user=V(topu1,topu2,...,topun) (2)
wherein, V (top)un) Representative sequence number topunA corresponding commodity room;
s300: constructing a collaborative filtering model based on a commodity house, and comprising the following steps;
s310: ranking UserScore according to similarity scores of historical payersnThe public deposit historical payer corresponding to the highest score in the group is used as the historical payer most similar to the to-be-detected payer, and the commodity house purchased by the most similar historical payer is used as the initial push commodity house;
s320: calculating similarity scores between the initial pushed commodity house and the rest m-1 historical commodity houses;
s330: the similarity scores are arranged according to the scores in a descending order to obtain the commodity house similarity score ranking HouseScorenAnd a list of commercial houses Rec _ house, House corenThe similarity scores in the Rec _ house are in one-to-one correspondence with the commodity rooms in the Rec _ house, and the similarity scores are ranked as HouseStorenAs shown in formula (3), the commodity room list Rec _ house is shown in formula (4):
HouseScorem=Sim(toph1,toph2,...,tophm) (3)
wherein, tophmIndicating a sequence number, Sim (top) ranked in descending order of the magnitude of the similarity scorehm) Denotes tophmA corresponding similarity score;
Rec_house=V(toph1,toph2,...,tophm) (4)
wherein, V (top)hm) Representative sequence number tophmA corresponding commodity room;
s400: acquiring commodity house information ghouse which can be purchased by using a public accumulation fund loan;
s500: and carrying out deduplication processing on Rec _ user and Rec _ house to obtain a commodity room list Rec _ temp after deduplication, wherein the expression is as follows:
Rec_temp=Rec_user∪Rec_house (5);
s600: intersecting the Rec _ temp and the ghouse to obtain a commodity room list Rec _ list which is finally pushed to a to-be-detected payer, wherein the specific calculation expression is as follows:
Rec_list=ghouse∩Rec_temp。
2. the method for matching the house purchasing precision of the public accumulation fund payment people based on the double collaborative filtering as claimed in claim 1, wherein: the specific method for calculating the similarity score between the to-be-detected payer information and each historical payer information in S230 is a Jaccard (Jaccard) similarity coefficient method, and the specific expression is as follows:
3. The method for matching the house purchasing precision of the public accumulation fund payment people based on the double collaborative filtering as claimed in claim 2, wherein: the similarity calculation in S320 uses a cosine similarity calculation method, and the specific expression is as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210129340.9A CN114461926A (en) | 2022-02-11 | 2022-02-11 | Precise matching method for public accumulation fund payment person house purchasing based on double collaborative filtering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210129340.9A CN114461926A (en) | 2022-02-11 | 2022-02-11 | Precise matching method for public accumulation fund payment person house purchasing based on double collaborative filtering |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114461926A true CN114461926A (en) | 2022-05-10 |
Family
ID=81412974
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210129340.9A Pending CN114461926A (en) | 2022-02-11 | 2022-02-11 | Precise matching method for public accumulation fund payment person house purchasing based on double collaborative filtering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114461926A (en) |
-
2022
- 2022-02-11 CN CN202210129340.9A patent/CN114461926A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108647996B (en) | Spark-based personalized recommendation method and system | |
CN105224699B (en) | News recommendation method and device | |
Christensen et al. | Social group recommendation in the tourism domain | |
CN103679462B (en) | A kind of comment data treating method and apparatus, a kind of searching method and system | |
CN108805598B (en) | Similarity information determination method, server and computer-readable storage medium | |
CN109918563B (en) | Book recommendation method based on public data | |
CN110110225B (en) | Online education recommendation model based on user behavior data analysis and construction method | |
CN101454771A (en) | System and method of segmenting and tagging entities based on profile matching using a multi-media survey | |
CN106528812B (en) | A kind of cloud recommended method based on USDR model | |
CN111611499B (en) | Collaborative filtering method, collaborative filtering device and collaborative filtering system | |
CN111026977B (en) | Information recommendation method and device and storage medium | |
US20190377905A1 (en) | Mixed Data Fingerprinting with Principal Components Analysis | |
CN104050243A (en) | Network searching method and system combined with searching and social contact | |
CN114266443A (en) | Data evaluation method and device, electronic equipment and storage medium | |
CN114036376A (en) | Time-aware self-adaptive interest point recommendation method based on K-means clustering | |
CN111429161B (en) | Feature extraction method, feature extraction device, storage medium and electronic equipment | |
CN114528480A (en) | Time-sensing self-adaptive interest point recommendation method based on K-means clustering | |
CN115422441A (en) | Continuous interest point recommendation method based on social space-time information and user preference | |
CN111475744A (en) | Personalized position recommendation method based on ensemble learning | |
CN111144831A (en) | Accurate selection screening system and method suitable for recruitment | |
CN110543601B (en) | Method and system for recommending context-aware interest points based on intelligent set | |
Bhuiyan et al. | Waiting to be sold: Prediction of time-dependent house selling probability | |
CN109992592B (en) | College poverty and poverty identification method based on flow data of campus consumption card | |
CN109299368B (en) | Method and system for intelligent and personalized recommendation of environmental information resources AI | |
CN116010696A (en) | News recommendation method, system and medium integrating knowledge graph and long-term interest of user |
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 |