CN111159341B - Information recommendation method and device based on user investment and financial management preference - Google Patents
Information recommendation method and device based on user investment and financial management preference Download PDFInfo
- Publication number
- CN111159341B CN111159341B CN201911355303.4A CN201911355303A CN111159341B CN 111159341 B CN111159341 B CN 111159341B CN 201911355303 A CN201911355303 A CN 201911355303A CN 111159341 B CN111159341 B CN 111159341B
- Authority
- CN
- China
- Prior art keywords
- investment
- data
- information
- financing
- generating
- 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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an information recommendation method and device based on user investment and financing preference, wherein the method comprises the following steps: generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user; generating an article label according to the investment and financing preference data by using a domain knowledge map method and a natural language processing method; generating a matching retrieval rule according to the article label; recommending information to the user according to the matching search rule. The method leads the information recommendation to be capable of further understanding the investment and financing related products and the related information thereof by introducing the product relation knowledge map.
Description
Technical Field
The invention relates to the technical field of financial information, in particular to an information recommendation method and device based on user investment and financial management preference.
Background
With the development of the times, the demands of clients on investment and financial management are increasing day by day, and the professional demands of the clients are stronger and stronger. In the era of intelligent finance, value-added services such as information services are becoming more and more important in the aspects of customer education, customer stickiness maintenance and the like. Information explosion in the internet era, on one hand, various information is various and huge in quantity; on the other hand, the client is fragmented in time, and cannot quickly search out the information which is interested in the client from a large amount of information, particularly aiming at the field with very strong specialty of the investment and financing field. Although a large number of content service platforms such as today's headlines, internet news and the like exist at present, the content service platforms cannot be deeply and carefully analyzed in the investment and financing field to obtain a good recommendation effect, and most of the content service platforms are based on group intelligence analysis such as collaborative filtering and the like.
The information service is more and more valued as a value-added service in the banking services, and the value thereof is more and more increased. Some traditional recommendation models and recommendation processes cannot well utilize the existing rich financial data, including transaction, position taking and the like data generated by customers in various product systems. Many existing recommendation systems do not provide deep analysis and modeling for the financial investment and financing field. Aiming at the field of investment and financial management, if the specialty and the accuracy of the information recommendation service are improved, the problem to be solved urgently is solved. Most of products providing information services in the investment and financing field in the current market are recommended according to the topic popularity and the novelty, and personalized recommendation is not considered too much. The popular content service applications such as today's headlines, internet news and the like mainly build reading interests of the clients based on the text contents of historical reading articles of the clients, and do not combine with related data of investment and financial management of the clients.
Disclosure of Invention
Aiming at the problems in the prior art, the information recommendation method based on the user investment and financing preference builds a set of perfect investment and financing field labels and realizes the recommendation of relevant information from the aspects of content and relevant products for information articles. The method supports mass information data, and has the advantages of high query efficiency, high recommendation accuracy and low implementation cost.
In order to solve the technical problems, the invention provides the following technical scheme:
in a first aspect, the present invention provides an information recommendation method based on user investment financing preferences, comprising:
generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user;
generating article labels according to the investment financing preference data by using a domain knowledge map method and a natural language processing method;
generating a matching retrieval rule according to the article label;
recommending information to the user according to the matching search rule.
In one embodiment, the historical investment financing data comprises: the investment financing transaction data and the investment financing behavior data comprise the display, click, like, share, collection and shielding of various information.
In one embodiment, the generating investment financing preference data according to the historical investment financing data, the behavior data and the current position taken data of the user comprises:
generating an interest model of the user according to the behavior data;
and generating the investment and financing preference data according to the interest model and the position data.
In one embodiment, the generating an interest model of the user according to the behavior data includes:
respectively generating a long-term interest model and a short-term interest model according to preset time and the behavior data by using a streaming processing method;
and updating the short-term interest model in real time.
In an embodiment, the generating a matching retrieval rule according to the article tag includes:
sorting the information corresponding to the article label by using Wide & Deep and Deep learning models;
the information similarity is calculated by cosine similarity or correlation coefficient method.
In a second aspect, the present invention provides an information recommendation apparatus based on user investment financing preferences, the apparatus comprising:
the investment financing preference data generation unit is used for generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user;
the article label generating unit generates article labels according to the investment financing preference data by using a domain knowledge map method and a natural language processing method;
the matching retrieval rule generating unit is used for generating a matching retrieval rule according to the article label;
and the information recommending unit is used for recommending information to the user according to the matching search rule.
In one embodiment, the historical investment financing data comprises: the investment financing transaction data and the investment financing behavior data comprise the display, click, like, share, collection and shielding of various information.
In one embodiment, the investment financing preference data generating unit includes:
the interest model generating module is used for generating an interest model of the user according to the behavior data;
and the investment financing preference data generation module is used for generating the investment financing preference data according to the interest model and the position data.
In one embodiment, the interest model generation module comprises:
the interest model generation module is used for respectively generating a long-term interest model and a short-term interest model according to preset time and the behavior data by using a streaming processing method;
and the interest model updating module is used for updating the short-term interest model in real time.
In one embodiment, the matching search rule generating unit includes:
the information sequencing module is used for sequencing the information corresponding to the article label by using Wide & Deep and Deep learning models;
and the similarity calculation module is used for calculating the information similarity by using a cosine similarity or correlation coefficient method.
In a third aspect, the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the information recommendation method based on the user investment and financial management preference.
In a fourth aspect, the present invention provides a computer readable storage medium, having stored thereon a computer program, which, when being executed by a processor, performs the steps of a method for information recommendation based on user investment financing preferences.
From the above description, the embodiment of the invention provides an information recommendation method and device based on user investment financing preference, aiming at information article information in investment financing service, analyzing the client investment financing preference by using position data and transaction detail data of clients in various investment financing product systems of a bank, introducing semantic extension information by combining with an investment financing domain knowledge map, constructing an investment financing domain label for the information article, and performing matching mapping with the client investment financing preference, thereby realizing a set of information recommendation service method based on client investment financing preference. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the investment financing field.
3. The training data of the sorting and sequencing model is combined with the investment and financing related data of the client, and the investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the recall layer retrieval keywords.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-term and short-term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation strategy manager to uniformly manage and configure different recommendation strategies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark and streaming computing capability is provided by combination of Kafka, so that client behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and associated information thereof.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic flow chart illustrating an information recommendation method based on user investment financing preference in an embodiment of the present invention;
FIG. 2 is a flowchart illustrating the steps 100 of the information recommendation method based on the user's investment financing preference according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating the step 1011 of the information recommendation method based on the user investment financing preference according to the embodiment of the present invention;
FIG. 4 is a flowchart illustrating the steps 300 of the information recommendation method based on the user's investment financing preference according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart illustrating a method for recommending information based on user investment financing preferences in an embodiment of the present invention;
FIG. 6 is a schematic diagram illustrating a process for modeling interest of a client in an embodiment of the present invention;
FIG. 7 is a diagram illustrating a method for recommending an architecture based on client investment preference information according to an embodiment of the present invention;
FIG. 8 is a flowchart of online recommendation in an embodiment of the present invention;
FIG. 9 is a flow chart of a recommendation policy manager in an embodiment of the present invention;
FIG. 10 is a flow chart illustrating the process of the article recall layer in accordance with an embodiment of the present invention;
FIG. 11 is a flow chart illustrating a process of an information article ranking layer in an embodiment of the present invention;
FIG. 12 is a flow chart illustrating tag processing of an information article according to an embodiment of the present invention;
FIG. 13 is a flowchart illustrating a related information calculation process according to an embodiment of the present invention;
FIG. 14 is a schematic diagram illustrating a process for sorting training data of a ranking model in an embodiment of the present invention;
FIG. 15 is a block diagram of an information recommendation apparatus based on user investment financing preferences in an embodiment of the present invention;
fig. 16 is a block diagram of an investment financing preference data generating unit in the embodiment of the present invention;
FIG. 17 is a block diagram of an interest model generation module in an embodiment of the invention;
fig. 18 is a block diagram of a matching search rule generation unit in an embodiment of the present invention;
fig. 19 is a schematic structural diagram of an electronic device in an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Based on the shortcomings of the prior art, the embodiment of the present invention provides a specific implementation of an information recommendation method based on user investment and financial management preferences, referring to fig. 1, the method specifically includes the following steps:
step 100: and generating investment financing preference data according to the historical investment financing data, the behavior data and the current position data of the user.
It is understood that the historical investment financing data in step 100 includes: the investment financing transaction data comprises investment financing transaction data and investment financing behavior data, wherein the investment financing behavior data comprises the display, click, like, share, collection and shielding of various information.
Step 200: and generating article labels according to the investment financing preference data by using a domain knowledge map method and a natural language processing method.
Specifically, a label system is constructed by combining investment financing business and product characteristics, and different financial markets and product types are considered, wherein the label system mainly comprises a fund market, a stock market, a trust market, a bond market, a precious metal market and the like. Topics, entities and the like related to various markets are used as secondary or tertiary labels to construct a complete label system. And labeling the information article by combining a multi-label classification model processed by natural language and heuristic rules.
Step 300: and generating a matching retrieval rule according to the article label.
Firstly, semantic retrieval is carried out in a product knowledge graph according to the subject labels of the information articles and the entity mentions appearing in the article contents, retrieval results can be directly matched with certain products or certain investment field subjects, investment financing subject vectors representing the information articles are extracted, and then the similarity of different information in the investment financing subject direction is calculated according to the obtained investment financing subject vectors.
Step 400: recommending information to the user according to the matching search rule.
From the above description, the embodiment of the invention provides an information recommendation method based on user investment and financial management preferences, aiming at information article information in investment and financial management services, the position data and transaction detail data of a client in various investment and financial management product systems of a bank are utilized to analyze the client investment and financial management preferences, semantic extension information is introduced by combining with an investment and financial management domain knowledge map, an investment and financial management domain label is constructed for the information article, and the information article is matched and mapped with the client investment and financial management preferences, so that a set of complete information recommendation service method based on the client investment and financial management preferences is realized.
In one embodiment, referring to fig. 2, step 100 specifically includes:
step 101: and generating an interest model of the user according to the behavior data.
It is understood that the customer reading interests are divided into long-term interests and short-term interests. As the time of the customer using the information service system goes on, a large amount of customer behavior data can be accumulated, which mainly comprises behavior data of information display, click-to-read data, like, dislike, collection, sharing, shielding and the like. Preferably, the construction of long-term interest is based on historical data over a long time horizon. The construction of the short-term interest is based on a streaming processing mode, kafka accesses information click reading data, and SparkStream subscribes to each click reading record consumed by Kafka in real time. When the user accesses an information click reading data, the relevant information labels and the weights of the relevant information labels of the user are updated in time, and the label weight of the recently read article is higher.
Step 102: and generating the investment and financing preference data according to the interest model and the position data.
In one embodiment, referring to fig. 3, step 101 further comprises:
step 1011: and respectively generating a long-term interest model and a short-term interest model according to preset time and the behavior data by using a streaming processing method.
Step 1012: and updating the short-term interest model in real time.
And constructing a client interest model based on the showing, clicking and behavior data such as praise, like, forwarding, shielding and the like generated by the client aiming at the information. Preferably, the client short-term interest model is updated based on streaming computing using SparkStream to process client behavior data in real-time. And constructing a long-term interest model of the client based on the behavior data in the long period range, wherein the updating frequency of the long-term interest model is low, and the updating frequency is half a month or one month generally.
In one embodiment, referring to fig. 4, step 300 specifically includes:
step 301: and sequencing the information corresponding to the article label by using Wide & Deep and Deep learning models.
It is understood that the Wide & Deep and Deep learning models of Deep and Deep Cross are modeled using a Keras based on the tenserflow back end. And collecting the display data and the click data in a certain time period as training data for model training, wherein the time range of data collection can be adjusted at will.
Step 302: the information similarity is calculated by cosine similarity or correlation coefficient method.
The conventional way to calculate the related information is to calculate the literal similarity based on the text content, such as using cosine similarity. The method has higher professional degree and business purpose for the calculation of the related information in the investment and financing field. Firstly, semantic retrieval is carried out in a product knowledge graph according to the subject labels of the information articles and the entity mentions appearing in the article contents, retrieval results can be directly matched with certain products or certain investment field subjects, investment financing subject vectors representing the information articles are extracted, and then the similarity of different information in the investment financing subject direction is calculated according to the obtained investment financing subject vectors. The similarity calculation can be performed by methods such as cosine similarity or correlation coefficient.
From the above description, the embodiment of the invention provides an information recommendation method based on user investment and financial management preferences, aiming at information article information in investment and financial management services, the position data and transaction detail data of a client in various investment and financial management product systems of a bank are utilized to analyze the client investment and financial management preferences, semantic extension information is introduced by combining with an investment and financial management domain knowledge map, an investment and financial management domain label is constructed for the information article, and the information article is matched and mapped with the client investment and financial management preferences, so that a set of complete information recommendation service method based on the client investment and financial management preferences is realized. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the investment financing field.
3. Training data of the sorting and ordering model is combined with investment and financing related data of the client, and investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the recall layer retrieval keywords.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-term and short-term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation strategy manager to uniformly manage and configure different recommendation strategies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark and streaming computing capability is provided by combination of Kafka, so that client behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and related information thereof.
To further illustrate the present solution, the present invention provides a specific application example of the information recommendation method based on the user investment financing preference, and the specific application example specifically includes the following contents, see fig. 5.
The specific application example constructs an asset relationship map of a client by accessing position taking and transaction data of product systems such as agency fund, precious metal, foreign exchange and the like. HBase and Hive are used as main storage, solr is used as a search engine, and Spark is used as a calculation engine.
S0: and generating an interest model of the user according to the behavior data.
On one hand, the information of products held, concerned or traded by the client can be directly obtained based on the asset relationship; and on the other hand, a client interest model is constructed based on the showing, clicking and behavior data such as praise, liking, forwarding, shielding and the like generated by the client in the information service system. And processing the client behavior data in real time by utilizing the spark stream based on streaming calculation, and updating the client short-term interest model. And constructing a long-term interest model of the client based on the behavior data in the long period range, wherein the updating frequency of the long-term interest model is low, and the updating frequency is half a month or one month generally. A perfect label system is built from the perspective of investment and financing for the information articles, and the function of recommending related information is realized for the subject of the articles, related products and the like. And semantic expansion is performed by means of the knowledge graph to improve the diversity and accuracy of information recommendation, see fig. 6.
S1: and generating the investment and financing preference data according to the interest model and the position data.
S2: and generating an article label according to the investment and financing preference data by utilizing a domain knowledge map method and a natural language processing method.
Specifically, a set of complete label system is established for the information articles in the investment field, and multi-label classification processing is carried out on the accessed information articles. The long-term interest of the client is obtained according to the subject and the contained label of the historical reading article (the historical data of the long time range) of the client, for example, the distribution of the article subject read by the client in one month is fund (53%), stock (23%), noble metal (15%) and others (9%), then the long-term interest of the client can be determined as fund, stock and noble metal, and the interest weights are fund (0.53/(0.53 +0.23+ 0.15) = 0.58), stock (0.23/(0.53 +0.23+ 0.15) = 0.25) and noble metal (0.15/(0.53 +0.23+ 0.15) = 0.17), respectively. The short-term interest of the client is mainly used for tracking the current instant reading interest of the client and is determined by the theme and the label corresponding to the article which is read by the client through recent clicking. The long-term interest is modeled by the reading history of the client in the last 1 month or 3 months, and the short-term interest is determined according to the article read by the client in the last N clicks (N is suggested to be 1-10).
In addition, aiming at articles, related information calculation is realized from different angles of titles, texts, themes, labels and content mentions (related products), and the requirements of recommendation of different related information are met. And introducing a knowledge graph based on various product relation data to semantically expand the search terms in the information article recall step. The scheme is mainly divided into an online recommendation part and an offline batch processing part, and is shown in figure 7.
The online recommendation method comprises the steps of firstly identifying a client and a requester through a policy manager to obtain a recommendation policy configured by the requester. The whole recommendation process is divided into three parts, namely recall, accurate sequencing and list optimization, and requirements of accuracy, diversity, novelty and the like of recommendation results are fully considered. The flow of online recommendation is shown in fig. 8.
S3: the client initiates the request.
The client request mainly comprises a client unique identification number and a requester number. The unique client identification number is used for identifying the client in the information recommendation system, and the requester number is used for identifying the requester in the recommendation process.
S4: a policy manager is generated.
Referring to FIG. 9, a policy manager maintains a unified management of recommended policies configured by requesters for different types of customers. As a unified information recommendation platform, it is required to be flexibly compatible with different recommendation requirements. The policy manager includes two parts, a rule set and a rule engine. Wherein the basic rules are uniformly customized by a business expert. A service management desk is also provided for service personnel to configure specific rules. The rules are mainly configured from: group differences, branch differences, scene differences, channel differences, and information factor differences.
Each recommendation strategy can be configured according to various attributes with different dimensions and granularities, such as sources, themes, genres, release time, related investment products and the like of the information articles. The information related to the introduction of the large deposit list or other deposit products can be configured and recommended according to different scenes, such as different time points, for example, the end of the year. The recommendation rules can also be configured according to different customer groups, for example, for a high-net-value customer group, information related to preferentially recommending interest products can be configured. The branch difference is also the difference of the organization, different branches or branches can be configured with recommendation rules suitable for themselves, and the rules can be suitable for customers under the organization. The channel difference refers to different recommendation rules configured for different channels of information service deployment, for example, it is not suitable for an intelligent terminal such as STM to push information with very long content, and it is suitable for short information or video information. The rule engine calculates an adaptive rule set according to request information such as client information, transaction information, organization information, channel information and the like of the sending request, and forms a final recommendation strategy according to the obtained rules. Each rule corresponds to a recall policy in the recommendation process. For example, if an organization currently sets the highest priority rule as "recommend xxx subject information", the information retrieved by the subject field equal to xxx condition will be prioritized during the recall process.
The recommendation strategy can be configured from two aspects of recommendation and non-recommendation (shielding), and various recommendation requirements can be met to the maximum extent.
S5: generate the recall layer of the information article.
Referring to FIG. 10, the information article recall layer screens out a relatively small number of recommended candidate articles from a large number of information articles according to different strategies and perspectives. The recall layer relies on the recommended policies obtained by the policy manager as a screening basis. And semantically expanding the retrieval conditions based on the client investment and financing preference information and reading interest and in combination with a product knowledge graph. For example, a customer holding a fund product is found, and the multi-dimensional information of the product, such as the category information, fund level, administrator information, the fund company and relevant attributes thereof, can be obtained from the product knowledge graph as an extended retrieval condition. The policy manager is responsible for directing and constraining all processing details of the recall layer.
S6: and sequencing the information corresponding to the article label by using Wide & Deep and Deep learning models.
Referring to FIG. 11, the number of candidate recommended article collections available at the recall level is still relatively large, typically 10-20 pieces of information per request from the client. The results returned for the recall layer also need to be sorted accurately. The project mainly uses two Deep learning models, namely Wide & Deep and Deep & Cross, as the ranking models. These two models were modeled using Keras based on the tensrflow back end. And collecting the display data and the click data in a certain time period as training data for model training, wherein the time range of data collection can be adjusted at will. Taking records which are recommended to the client but are not clicked and viewed by the client as a negative sample; the recommendation is given to the client, and the client clicks on the viewed record as a sample. The selected characteristics mainly comprise the following parts:
(1) A customer portrait label: the method is mainly generated from transaction data and asset data of a customer in a bank, and is pre-generated for a building customer. Such as investment product preferences, whether it is a moonlight family, risk tolerance, etc.
(2) The client reads the interest tag: the article tags clicked on by the client in a certain time period (the tags of the clicked articles) are collected. The characteristic value is the click rate of each label by the client.
(3) The basic information of the client: including the customer's sex, age, address, etc
(4) Customer behavior data: the online proportion of different time periods, the average online time length of a single day, the average reading time length of a single seal and the like
(5) Customer transaction information: traded investment financing products, average traded amount, average time to stay, etc.
(6) Information elements: including source of information, time of release, subject, label, length of content, type, etc.
(7) Information history statistical information: information click times, collection times, praise times, sharing times, etc.
The Google has been widely applied in the recommendation field after proposing the two models, but data and attributes related to client investment and financing are mainly considered as characteristics for the training data book project. The massive data generated by the customers in the product transaction system is fully utilized. The training data combined with the client investment and financial management data is utilized to enable the model to automatically mine and analyze the relation between the client investment and financial management preference and the article, so that the final recommendation result is more accurate.
It will be appreciated that the list optimization further adjusts the final ranking results primarily based on other requirements of the recommendation system, such as considering interest heuristics, and recommending information articles to the client that the client does not often see or that the client rarely reads.
S7: and (4) performing offline batch processing.
Compared with online processing, the offline batch processing is mainly responsible for labeling information articles, constructing reverse and forward indexes, modeling client interests (long and short term interests), calculating related information, sorting training data of a model and the like. A set of perfect label system is constructed by combining investment financing business and product characteristics, and different financial markets and product types are considered, wherein the label system mainly comprises a fund market, a stock market, a trust market, a bond market, a precious metal market and the like. Topics, entities and the like related to various markets are used as secondary or tertiary labels to construct a complete label system. The information article is labeled by combining a multi-label classification model of natural language processing and heuristic rules, and the method is shown in figure 12.
S8: the information similarity is calculated by cosine similarity or correlation coefficient method.
Referring to fig. 13, the conventional method for calculating related information is to calculate literal similarity based on text content, such as cosine similarity. The method has higher professional degree and business purpose for the calculation of the related information in the investment and financing field. Firstly, semantic retrieval is carried out in a product knowledge graph according to the theme labels of the information articles and entity mentions appearing in article contents, retrieval results can be directly matched with themes of certain products or certain investment fields, investment financing theme vectors representing the information articles are extracted, and then the similarity of different information in the investment financing theme direction is calculated according to the obtained investment financing theme vectors. The similarity calculation can be performed by methods such as cosine similarity or correlation coefficient. Secondly, from the perspective of business, the system provides a function of inputting a heavily recommended label, business personnel can add a new label and configure an article entity mention corresponding to the new label. The articles that appear to be referenced by the configured entity increase their relevance weight.
S9: and (4) collecting and sorting training data.
Over time, the information service system accumulates a great deal of customer behavior data, but training the ranking model by only relying on the behavior data does not take into account the investment and financial preference of the customer. Therefore, the project considers the basic information of the customers, the historical behavior data of the customers and the detailed data of the customers in each product transaction system when collecting the training data of the ranking model. The final training data was obtained by integrating these data, see fig. 14.
From the above description, the embodiment of the invention provides an information recommendation method based on user investment and financial management preferences, aiming at information article information in investment and financial management services, the position data and transaction detail data of a client in various investment and financial management product systems of a bank are utilized to analyze the client investment and financial management preferences, semantic extension information is introduced by combining with an investment and financial management domain knowledge map, an investment and financial management domain label is constructed for the information article, and the information article is matched and mapped with the client investment and financial management preferences, so that a set of complete information recommendation service method based on the client investment and financial management preferences is realized. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the field of investment and financing.
3. Training data of the sorting and ordering model is combined with investment and financing related data of the client, and investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the retrieval keywords of the recall layer.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-short term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation strategy manager to uniformly manage and configure different recommendation strategies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark, streaming computing capability is provided by Kafka combination, and customer behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and associated information thereof.
Based on the same inventive concept, the embodiment of the present application further provides an information recommendation device based on the user investment and financing preference, which can be used to implement the method described in the above embodiment, as described in the following embodiments. Because the principle of solving the problems of the information recommendation device based on the user investment and financial management preference is similar to the information recommendation method based on the user investment and financial management preference, the implementation of the information recommendation device based on the user investment and financial management preference can refer to the implementation of the information recommendation method based on the user investment and financial management preference, and repeated parts are not repeated. As used hereinafter, the term "unit" or "module" may be a combination of software and/or hardware that implements a predetermined function. While the system described in the embodiments below is preferably implemented in software, implementations in hardware, or a combination of software and hardware are also possible and contemplated.
The embodiment of the invention provides a specific implementation mode of an information recommendation device based on user investment and financial management preference, which can realize an information recommendation method based on user investment and financial management preference, and referring to fig. 15, the information recommendation device based on user investment and financial management preference specifically comprises the following contents:
the investment financing preference data generation unit 10 is used for generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user;
an article label generating unit 20, which generates article labels according to the investment financing preference data by using a domain knowledge map method and a natural language processing method;
a matching search rule generating unit 30, configured to generate a matching search rule according to the article tag;
an information recommending unit 40 for recommending information to the user according to the matching search rule.
In one embodiment, the historical investment financing data comprises: the investment financing transaction data and the investment financing behavior data comprise the display, click, like, share, collection and shielding of various information.
In one embodiment, referring to fig. 16, the investment financing preference data generating unit 10 includes:
an interest model generation module 101, configured to generate an interest model of the user according to the behavior data;
an investment financing preference data generation module 102, configured to generate the investment financing preference data according to the interest model and the position data.
In one embodiment, referring to fig. 17, the interest model generating module 101 includes:
an interest model generating module 1011, configured to generate a long-term interest model and a short-term interest model according to preset time and the behavior data by using a streaming processing method;
an interest model updating module 1012 for updating the short-term interest model in real time.
In one embodiment, referring to fig. 18, the matching retrieval rule generating unit 30 includes:
the information ordering module 301 is configured to order the information corresponding to the article tag by using Wide & Deep and Deep learning models;
the similarity calculation module 302 is configured to calculate the information similarity by using a cosine similarity or correlation coefficient method.
From the above description, the embodiment of the invention provides an information recommendation device based on user investment and financial management preferences, aiming at information article information in investment and financial management services, the position data and transaction detail data of a client in various investment and financial management product systems of a bank are utilized to analyze the client investment and financial management preferences, semantic extension information is introduced by combining with an investment and financial management domain knowledge map, an investment and financial management domain label is constructed for the information article, and the information article is matched and mapped with the client investment and financial management preferences, so that a set of complete information recommendation service method based on the client investment and financial management preferences is realized. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the investment financing field.
3. The training data of the sorting and sequencing model is combined with the investment and financing related data of the client, and the investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the retrieval keywords of the recall layer.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-term and short-term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation strategy manager to uniformly manage and configure different recommendation strategies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of the related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark and streaming computing capability is provided by combination of Kafka, so that client behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and associated information thereof.
An embodiment of the present application further provides a specific implementation manner of an electronic device, which is capable of implementing all steps in the information recommendation method based on the user investment and financial management preference in the foregoing embodiment, and referring to fig. 19, the electronic device specifically includes the following contents:
a processor (processor) 1201, a memory (memory) 1202, a communication Interface (Communications Interface) 1203, and a bus 1204;
the processor 1201, the memory 1202 and the communication interface 1203 complete communication with each other through the bus 1204; the communication interface 1203 is configured to implement information transmission between related devices, such as a server-side device, a data acquisition device, and a client device.
The processor 1201 is configured to invoke the computer program in the memory 1202, and the processor implements all the steps of the information recommendation method based on the user investment financing preference in the above embodiments when executing the computer program, for example, the processor implements the following steps when executing the computer program:
step 100: and generating investment financing preference data according to the historical investment financing data, the behavior data and the current position data of the user.
Step 200: and generating an article label according to the investment and financing preference data by utilizing a domain knowledge map method and a natural language processing method.
Step 300: and generating a matching retrieval rule according to the article label.
Step 400: recommending information to the user according to the matching search rule.
From the above description, it can be seen that, in the electronic device in the embodiment of the present application, for information article information in investment financing service, the position data and transaction detail data of a client in various investment financing product systems of a bank are utilized to analyze client investment financing preference, semantic extension information is introduced in combination with a knowledge map of the investment financing field, an investment financing field label is constructed for the information article, and the information article is mapped in a matching manner with the client investment financing preference, so that a set of complete information recommendation service method based on the client investment financing preference is implemented. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the field of investment and financing.
3. The training data of the sorting and sequencing model is combined with the investment and financing related data of the client, and the investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the recall layer retrieval keywords.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-short term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation policy manager to uniformly manage and configure different recommendation policies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark and streaming computing capability is provided by combination of Kafka, so that client behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and associated information thereof.
An embodiment of the present application further provides a computer-readable storage medium capable of implementing all the steps in the information recommendation method based on the user investment and financial management preference in the foregoing embodiment, where the computer-readable storage medium stores a computer program, and the computer program, when executed by a processor, implements all the steps in the information recommendation method based on the user investment and financial management preference in the foregoing embodiment, for example, when the processor executes the computer program, the processor implements the following steps:
step 100: and generating investment financing preference data according to the historical investment financing data, the behavior data and the current position data of the user.
Step 200: and generating an article label according to the investment and financing preference data by utilizing a domain knowledge map method and a natural language processing method.
Step 300: and generating a matching retrieval rule according to the article label.
Step 400: recommending information to the user according to the matching search rule.
From the above description, the computer-readable storage medium in the embodiment of the present application analyzes the client investment and financing preference by using the position data and the transaction detail data of the client in various investment and financing product systems of the bank, introduces semantic extension information by combining the knowledge map of the investment and financing field, constructs an investment and financing field label for the information article, and performs matching mapping with the client investment and financing preference, thereby implementing a set of complete information recommendation service methods based on the client investment and financing preference. Specifically, the present invention has the following advantages:
1. and analyzing the client investment financing preference based on the client investment financing related data detail as the basis of information recommendation.
2. A set of perfect information article label system is established aiming at the investment financing field.
3. Training data of the sorting and ordering model is combined with investment and financing related data of the client, and investment and financing related attributes of the client are introduced as features.
4. And introducing a knowledge graph constructed based on various investment and financial products and relationship data thereof to perform semantic expansion on the recall layer retrieval keywords.
5. When the information article is labeled, the product relation knowledge graph is used for pre-labeling processing, so that the labeling accuracy is improved.
6. And constructing a long-term and short-term reading interest model of the client in the investment and financing field based on a streaming processing technology and a big data processing technology.
7. And constructing a recommendation strategy manager to uniformly manage and configure different recommendation strategies and flexibly changeable recommendation requirements.
In conclusion, the invention takes big data as a basic platform and constructs a client interest model aiming at the field of investment and financing. A set of perfect investment financing field labels are built for the information articles, and the recommendation of related information is realized for the information articles from the aspects of content and related products. Solr is used as a search engine, massive information data is supported, the query efficiency is high, the recommendation accuracy is high, the implementation cost is low, and efficient distributed retrieval service is provided based on Solr; distributed computing is provided based on Spark and streaming computing capability is provided by combination of Kafka, so that client behavior data can be mined and analyzed quickly and efficiently. By introducing the product relation knowledge graph, the recommendation system can more deeply understand investment and financing related products and associated information thereof.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The principle and the implementation mode of the invention are explained by applying specific embodiments in the invention, and the description of the embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. An information recommendation method based on user investment financing preference is characterized by comprising the following steps:
generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user;
generating an article label according to the investment and financing preference data by using a domain knowledge map method and a natural language processing method;
generating a matching retrieval rule according to the article label;
recommending information to the user according to the matching search rule;
the generating of the matching retrieval rule according to the article label comprises the following steps:
sorting the information corresponding to the article label by using Wide & Deep and Deep learning models;
the information similarity is calculated by cosine similarity or correlation coefficient method.
2. The information recommendation method of claim 1, wherein said historical investment financing data comprises: the investment financing transaction data comprises investment financing transaction data and investment financing behavior data, wherein the investment financing behavior data comprises the display, click, like, share, collection and shielding of various information.
3. The information recommendation method of claim 1, wherein said generating investment financing preference data based on historical investment financing data, behavioral data and current position data of the user comprises:
generating an interest model of the user according to the behavior data;
and generating the investment and financing preference data according to the interest model and the position data.
4. The information recommendation method of claim 3, wherein said generating an interest model of the user based on said behavior data comprises:
respectively generating a long-term interest model and a short-term interest model according to preset time and the behavior data by using a streaming processing method;
and updating the short-term interest model in real time.
5. An information recommendation device based on user investment financing preference, comprising:
the investment financing preference data generation unit is used for generating investment financing preference data according to historical investment financing data, behavior data and current position data of the user;
the article label generating unit generates article labels according to the investment and financing preference data by using a domain knowledge map method and a natural language processing method;
a matching retrieval rule generating unit, configured to generate a matching retrieval rule according to the article tag;
the information recommending unit is used for recommending information to the user according to the matching search rule;
the matching retrieval rule generation unit includes:
the information sequencing module is used for sequencing the information corresponding to the article label by using Wide & Deep and Deep learning models;
and the similarity calculation module is used for calculating the information similarity by using a cosine similarity or correlation coefficient method.
6. The information recommendation device of claim 5, wherein said historical investment financing data comprises: the investment financing transaction data comprises investment financing transaction data and investment financing behavior data, wherein the investment financing behavior data comprises the display, click, like, share, collection and shielding of various information.
7. The information recommendation apparatus according to claim 5, wherein said investment financing preference data generating unit comprises:
the interest model generating module is used for generating an interest model of the user according to the behavior data;
and the investment financing preference data generation module is used for generating the investment financing preference data according to the interest model and the position data.
8. The information recommendation device of claim 7, wherein the interest model generation module comprises:
the interest model generating module is used for respectively generating a long-term interest model and a short-term interest model according to preset time and the behavior data by utilizing a streaming processing method;
and the interest model updating module is used for updating the short-term interest model in real time.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the steps of the information recommendation method based on user investment and financial management preference according to any one of claims 1 to 4.
10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when being executed by a processor, implements the steps of the information recommendation method based on user investment financing preference according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911355303.4A CN111159341B (en) | 2019-12-25 | 2019-12-25 | Information recommendation method and device based on user investment and financial management preference |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911355303.4A CN111159341B (en) | 2019-12-25 | 2019-12-25 | Information recommendation method and device based on user investment and financial management preference |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111159341A CN111159341A (en) | 2020-05-15 |
CN111159341B true CN111159341B (en) | 2023-04-14 |
Family
ID=70558002
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911355303.4A Active CN111159341B (en) | 2019-12-25 | 2019-12-25 | Information recommendation method and device based on user investment and financial management preference |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111159341B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111680073A (en) * | 2020-06-11 | 2020-09-18 | 天元大数据信用管理有限公司 | Financial service platform policy information recommendation method based on user data |
CN111986006A (en) * | 2020-09-02 | 2020-11-24 | 深圳壹账通智能科技有限公司 | Product recommendation method and device based on knowledge graph, computer equipment and storage medium |
CN112102097A (en) * | 2020-09-28 | 2020-12-18 | 中国银行股份有限公司 | Investment capacity evaluation method and device, electronic equipment and computer storage medium |
CN113240472B (en) * | 2021-05-19 | 2024-02-02 | 深圳索信达数据技术有限公司 | Financial product recommendation method, electronic equipment and storage medium |
CN114463136A (en) * | 2022-01-06 | 2022-05-10 | 弘量科技(深圳)有限公司 | Fund recommendation method, apparatus and device based on multidimensional data mining and logic deduction |
CN116150477A (en) * | 2022-12-06 | 2023-05-23 | 上海贝耳塔信息技术有限公司 | Financial information personalized recommendation method, device, equipment and medium |
CN118069930A (en) * | 2024-04-17 | 2024-05-24 | 杭州高能云科技有限公司 | Investment dynamic recommendation method and system based on matching analysis |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107885886A (en) * | 2017-12-07 | 2018-04-06 | 百度在线网络技术(北京)有限公司 | To the method, apparatus and server of information recommendation sort result |
CN110245259A (en) * | 2019-05-21 | 2019-09-17 | 北京百度网讯科技有限公司 | The video of knowledge based map labels method and device, computer-readable medium |
CN110457581A (en) * | 2019-08-02 | 2019-11-15 | 达而观信息科技(上海)有限公司 | A kind of information recommended method, device, electronic equipment and storage medium |
CN110458641A (en) * | 2019-06-28 | 2019-11-15 | 苏宁云计算有限公司 | A kind of electric business recommended method and system |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150006294A1 (en) * | 2013-06-28 | 2015-01-01 | Linkedln Corporation | Targeting rules based on previous recommendations |
US20190129958A1 (en) * | 2017-10-30 | 2019-05-02 | Facebook, Inc. | Optimizing the Mapping of Qualitative Labels to Scores for Calculating Gain in Search Results |
-
2019
- 2019-12-25 CN CN201911355303.4A patent/CN111159341B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107885886A (en) * | 2017-12-07 | 2018-04-06 | 百度在线网络技术(北京)有限公司 | To the method, apparatus and server of information recommendation sort result |
CN110245259A (en) * | 2019-05-21 | 2019-09-17 | 北京百度网讯科技有限公司 | The video of knowledge based map labels method and device, computer-readable medium |
CN110458641A (en) * | 2019-06-28 | 2019-11-15 | 苏宁云计算有限公司 | A kind of electric business recommended method and system |
CN110457581A (en) * | 2019-08-02 | 2019-11-15 | 达而观信息科技(上海)有限公司 | A kind of information recommended method, device, electronic equipment and storage medium |
Non-Patent Citations (1)
Title |
---|
大规模知识图谱及其应用研究;孙雨生等;《情报理论与实践》;20181127(第11期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN111159341A (en) | 2020-05-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111159341B (en) | Information recommendation method and device based on user investment and financial management preference | |
US9704185B2 (en) | Product recommendation using sentiment and semantic analysis | |
JP3389948B2 (en) | Display ad selection system | |
CN108885624B (en) | Information recommendation system and method | |
CN110532479A (en) | A kind of information recommendation method, device and equipment | |
CN112785397A (en) | Product recommendation method, device and storage medium | |
WO2020088058A1 (en) | Information generating method and device | |
Cleger-Tamayo et al. | Top-N news recommendations in digital newspapers | |
JP2004164578A (en) | Method and apparatus for categorizing and presenting document of distributed database | |
US20180293678A1 (en) | Method and apparatus for the semi-autonomous management, analysis and distribution of intellectual property assets between various entities | |
CN110111167A (en) | A kind of method and apparatus of determining recommended | |
US20110004508A1 (en) | Method and system of generating guidance information | |
CN110413888A (en) | A kind of books recommended method and device | |
CN112116426A (en) | Method and device for pushing article information | |
CN110795613A (en) | Commodity searching method, device and system and electronic equipment | |
US20200073925A1 (en) | Method and system for generating a website from collected content | |
CN111859138B (en) | Searching method and device | |
CN117149804A (en) | Data processing method, device, electronic equipment and storage medium | |
CN109299368B (en) | Method and system for intelligent and personalized recommendation of environmental information resources AI | |
CN111753151A (en) | Service recommendation method based on internet user behaviors | |
CN116431895A (en) | Personalized recommendation method and system for safety production knowledge | |
CN115858815A (en) | Method for determining mapping information, advertisement recommendation method, device, equipment and medium | |
Madadipouya | A location-based movie recommender system using collaborative filtering | |
CN109408725B (en) | Method and apparatus for determining user interest | |
CN111460300A (en) | Network content pushing method and device and storage medium |
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 | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20220915 Address after: 25 Financial Street, Xicheng District, Beijing 100033 Applicant after: CHINA CONSTRUCTION BANK Corp. Address before: 25 Financial Street, Xicheng District, Beijing 100033 Applicant before: CHINA CONSTRUCTION BANK Corp. Applicant before: Jianxin Financial Science and Technology Co.,Ltd. |
|
TA01 | Transfer of patent application right | ||
GR01 | Patent grant | ||
GR01 | Patent grant |