CN113762392A - Financial product recommendation method, device, equipment and medium based on artificial intelligence - Google Patents
Financial product recommendation method, device, equipment and medium based on artificial intelligence Download PDFInfo
- Publication number
- CN113762392A CN113762392A CN202111052153.7A CN202111052153A CN113762392A CN 113762392 A CN113762392 A CN 113762392A CN 202111052153 A CN202111052153 A CN 202111052153A CN 113762392 A CN113762392 A CN 113762392A
- Authority
- CN
- China
- Prior art keywords
- product
- financial
- data
- financing
- historical
- 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
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 129
- 238000003066 decision tree Methods 0.000 claims abstract description 97
- 238000012216 screening Methods 0.000 claims abstract description 19
- 239000013598 vector Substances 0.000 claims description 47
- 238000004590 computer program Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 7
- 230000009193 crawling Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 10
- 238000013145 classification model Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 5
- 238000007405 data analysis Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 230000003542 behavioural effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 235000014347 soups Nutrition 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Bioinformatics & Computational Biology (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application belongs to the technical field of artificial intelligence and provides a financial product recommendation method, device, equipment and medium based on artificial intelligence, wherein the method comprises the following steps: constructing a corresponding relation between historical financing data and historical product preference information; according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, forming a training set, selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models; forming a financial product recommendation model by using the trained K decision tree models; the method comprises the steps of obtaining financing data of a user, inputting the financing data into a financing product recommendation model to obtain a plurality of financing products and product preference values, and screening out target financing products from the financing products to recommend the target financing products to the user, so that accurate recommendation is achieved, and the recommendation effect is good.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a financial product recommendation method, device, equipment and medium based on artificial intelligence.
Background
With the development of economy and the progress of science and technology, the financial consciousness of people is gradually enhanced, and more people invest idle funds into financial products. In order to promote own financial products to users, financial institutions such as banks and fund companies often choose users with strong willingness to purchase financial products to perform directional recommendation so as to improve the product recommendation effect.
In order to better understand the purchasing intention of the user, the financial platform usually evaluates the user according to the grasped user information, and recommends the financial product for the user with stronger intention of purchasing the financial product. However, when the current financial management platform recommends a user, only some popular financial management products are generally recommended, the recommended financial management products are too simple and are relatively fixed, and the recommendation effect is poor.
Disclosure of Invention
The application mainly aims to provide a financial product recommendation method, device, equipment and medium based on artificial intelligence so as to improve the recommendation effect of financial products.
In order to achieve the above object, the present application provides a financial product recommendation method based on artificial intelligence, which comprises the following steps:
acquiring historical financing data and historical product preference information of a user in a preset time period, and constructing a corresponding relation between the historical financing data and the historical product preference information;
according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, and taking all data sets and the historical product preference information corresponding to the data sets as training samples to form a training set;
selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
forming a financial product recommendation model by using the trained K decision tree models;
acquiring financial data of a user, and inputting the financial data into the financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product;
and screening out a target financing product from the financing products according to the product preference value, and recommending the target financing product to a user.
Preferably, the screening out the target financing product from the financing products according to the product preference value comprises:
acquiring current affair hotspot information, extracting a first keyword of the current affair hotspot information, and extracting a second keyword of each financial product;
calculating the similarity value of the first keyword and a second keyword of each financial product, and taking the second keyword with the similarity value larger than a preset similarity threshold value as a target keyword;
and taking the financing product corresponding to the target keyword as the target financing product.
Preferably, the first keyword includes a plurality of keywords, and the calculating of the similarity value of the first keyword and the second keyword of each financial product includes:
respectively converting the first keywords into Word vectors by using a Word2Vec Word vector model trained in advance, and calculating the average value of the Word vectors of the first keywords to obtain a first Word vector;
converting the second keyword of each financial product into a Word vector by using the Word2Vec Word vector model to obtain a second Word vector of each financial product;
and calculating the cosine distance between the first word vector and the second word vector of each financial product as the similarity value.
Preferably, the obtaining of the trained K decision tree models includes:
respectively calculating the loss value of each trained decision tree model by using a preset loss function;
judging whether the loss value of each decision tree model is lower than a preset loss value or not;
and if so, judging that the K decision tree models finish training.
Preferably, the screening out the target financing product from the financing products according to the product preference value comprises:
sorting the financial products in the order of the secondary magnitude of the product preference values;
taking the front N position of the financing products as target financing products; wherein N is a positive integer.
Preferably, the acquiring historical financial data of the user within the preset time period includes:
capturing a financing page browsed by the user within a preset time period through a crawling tool;
extracting information in the financing page by adopting a preset regular expression to obtain target information;
and analyzing the target information to obtain the historical financing data.
Preferably, the obtaining the historical financial data comprises:
dividing the analyzed target information into a plurality of sub-data blocks, and respectively allocating a unique identifier for each sub-data block according to a data structure;
and traversing all the identifiers, reserving a unique data block corresponding to each identifier, removing other data blocks with the same identifier, and obtaining the historical financial data with the repeated data removed.
The application also provides a financial product recommendation device based on artificial intelligence, and it includes:
the acquisition module is used for acquiring historical financing data and historical product preference information of a user in a preset time period and constructing a corresponding relation between the historical financing data and the historical product preference information;
the dividing module is used for dividing historical financing data with the same historical product preference information into the same data set according to the corresponding relation, and forming a training set by taking all data sets and the historical product preference information corresponding to the data sets as training samples;
the training module is used for selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
the composition module is used for forming a financial product recommendation model by utilizing the trained K decision tree models;
the input module is used for acquiring financial data of a user, inputting the financial data into the financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product;
and the recommending module is used for screening out a target financing product from the financing products according to the product preference value and recommending the target financing product to a user.
The present application further provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the steps of any of the above methods when executing the computer program.
The present application also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of any of the methods described above.
According to the financial product recommendation method, device, equipment and medium based on artificial intelligence, the corresponding relation between historical financial data and historical product preference information is established by acquiring the historical financial data and the historical product preference information of a user in a preset time period; according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, taking all data sets and historical product preference information corresponding to the data sets as training samples to form a training set, selecting K training samples from the training set, constructing K decision tree models, and inputting each training sample to different decision tree models for training to obtain K trained decision tree models; forming a financial product recommendation model by using the trained K decision tree models; acquiring financial data of a user, inputting the financial data into a financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product; and screening out target financing products from the financing products according to the product preference values, and recommending the target financing products to the user. The method comprises the steps of classifying historical financing data based on reflected historical product preference information, correspondingly training a decision tree model by utilizing each type of historical financing data, realizing independent training, improving training efficiency, and obtaining a financing product recommendation model by utilizing a combination of individually trained decision tree classification models, so that the generated financing product recommendation model has a better training result, and the recommendation precision of the financing product is improved; meanwhile, the financial products are screened based on the product preference values output by the financial product recommendation model, so that financial products suitable for users are screened out, accurate recommendation is achieved, and the recommendation effect is better.
Drawings
FIG. 1 is a schematic flow chart of a financial product recommendation method based on artificial intelligence according to an embodiment of the present application;
FIG. 2 is a block diagram schematically illustrating the structure of an artificial intelligence-based financial product recommendation apparatus according to an embodiment of the present application;
fig. 3 is a block diagram illustrating a structure of a computer device according to an embodiment of the present application.
The implementation, functional features and advantages of the objectives of the present application will be further explained with reference to the accompanying drawings.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.
The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
Referring to fig. 1, the application provides a financial product recommendation method based on artificial intelligence, which takes a server as an execution main body, where the server may be an independent server, or a cloud server that provides basic cloud computing services such as cloud service, a cloud database, cloud computing, a cloud function, cloud storage, Network service, cloud communication, middleware service, domain name service, security service, Content Delivery Network (CDN), and a big data and artificial intelligence platform.
In this application, the financial product recommendation method based on artificial intelligence is used for improving the recommendation effect of financial products, referring to fig. 1, in one embodiment, the financial product recommendation method based on artificial intelligence comprises the following steps:
s11, acquiring historical financing data and historical product preference information of a user in a preset time period, and constructing a corresponding relation between the historical financing data and the historical product preference information;
s12, dividing historical financing data with the same historical product preference information into the same data set according to the corresponding relation, and forming a training set by taking all data sets and the historical product preference information corresponding to the data sets as training samples;
s13, selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
s14, forming a financial product recommendation model by using the trained K decision tree models;
s15, acquiring financial data of a user, inputting the financial data into the financial product recommendation model to obtain a plurality of financial products and a product preference value corresponding to each financial product;
s16, screening out target financing products from the financing products according to the product preference values, and recommending the target financing products to users.
As described in the above step S11, the historical financial data may include age, income, marital status, academic history, asset level, type of financial product, risk level, investment term, amount of money to be invested, profitability, and the like. The historical product preference information includes behavioral data such as preferred investment funds, stocks or bonds, and the like.
The corresponding relation between the historical financing data and the historical product preference information is constructed for better dividing the historical financing data, for example, when the financing product type is that the occupation ratio of bonds is maximum in the financing products of the user based on the historical financing data analysis, the historical product preference information of the user is a favorite investment bond and belongs to a robust type, and the corresponding historical financing data is data related to the bond, including the investment amount, holding time and the like of the bond.
As described in step S12, the present embodiment classifies historical financial data based on the correspondence, divides historical financial data having the same historical product preference information into the same data set, and uses all data sets and the historical product preference information corresponding to the data sets as training samples to form a training set. For example, all data about bond investment belonging to favorite investment bonds are divided into a data set, the same data set and corresponding historical product preference information are used as a training sample, and a plurality of training samples are combined into a training set and stored.
As described in step S13, the present embodiment may use a machine learning model algorithm to perform recommendation matching on financial products to meet the core business objective. Before recommendation, a financial product recommendation model needs to be obtained through training. Specifically, the decision tree model can be used for carrying out classification learning on the training set, the rule that real customers buy the financial products and the corresponding product preference values are mined, the mapping relation between the real customers and the product preference values is established, and finally the corresponding financial product recommendation model is generated. For example, K training samples can be randomly extracted from a training set, K decision tree models are also constructed, each decision tree model inputs a unique training sample, the data quantity of historical financing data of each training sample is enough, multiple times of training are conducted on the corresponding decision tree models to obtain the trained decision tree models, each trained decision tree model can correspondingly identify a product preference information based on the financing data, a financing product recommendation model with a better recommendation effect is obtained through a separate training mode, training in the same time period can be achieved through an independent training mode, and training efficiency is improved.
As described in the above step S14, the K decision tree classification models may be randomly combined to obtain the trained financial product recommendation model, for example, the K decision tree classification models may be spliced to be combined into the trained financial product recommendation model, or any at least two of the K decision tree classification models may be selected to be spliced to be combined into the trained financial product recommendation model, so that the product recommendation model obtained by combining the individually trained decision tree classification models has a better training result, and the recommendation accuracy of the financial product is improved.
As described in the above steps S15-S16, the financial product recommendation model of the present application may further calculate a product preference value for each financial product based on the historical product preference information, the product preference value being used to evaluate the user' S preference for the financial product. Therefore, the financial product recommendation model can output the product preference value corresponding to each financial product besides the financial product needing to be recommended. Specifically, the financial data of a user needing to recommend a financial product is obtained, the financial data is input into a trained financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product, then a target financial product is screened out from the financial products according to the product preference values, and the target financial product is recommended to the user. For example, a financing product with the highest product preference value is screened out as a target financing product, and the target financing product is recommended to the user.
According to the financial product recommendation method based on artificial intelligence, the corresponding relation between historical financial data and historical product preference information is established by acquiring the historical financial data and the historical product preference information of a user in a preset time period; according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, taking all data sets and historical product preference information corresponding to the data sets as training samples to form a training set, selecting K training samples from the training set, constructing K decision tree models, and inputting each training sample to different decision tree models for training to obtain K trained decision tree models; forming a financial product recommendation model by using the trained K decision tree models; acquiring financial data of a user, inputting the financial data into a financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product; and screening out target financing products from the financing products according to the product preference values, and recommending the target financing products to the user. The method comprises the steps of classifying historical financing data based on reflected historical product preference information, correspondingly training a decision tree model by utilizing each type of historical financing data, realizing independent training, improving training efficiency, and obtaining a financing product recommendation model by utilizing a combination of individually trained decision tree classification models, so that the generated financing product recommendation model has a better training result, and the recommendation precision of the financing product is improved; meanwhile, the financial products are screened based on the product preference values output by the financial product recommendation model, so that financial products suitable for users are screened out, accurate recommendation is achieved, and the recommendation effect is better.
In an embodiment, in step S16, the screening out the target financing product from the financing products according to the product preference value may specifically include:
s161, acquiring current affair hotspot information, extracting a first keyword of the current affair hotspot information, and extracting a second keyword of each financial product;
s162, calculating the similarity value of the first keyword and the second keyword of each financial product, and taking the second keyword with the similarity value larger than a preset similarity threshold value as a target keyword;
and S163, taking the financing product corresponding to the target keyword as the target financing product.
S164, many financial products recommended to the user can be provided, and all financial products recommended to the user are obviously not reasonable, so that the terminal is required to combine hot subject matters, make weighted ranking for the hot products, and recommend the hot products to the user in pages according to scoring results.
Specifically, in this embodiment, the current affair hotspot information related to financing is acquired from each financing website, the current affair hotspot information is subjected to word segmentation to extract all keywords of the current affair hotspot information to obtain a first keyword, and description information of each financing product is extracted, the description information is a simple description of the financing product, then the description information is subjected to word segmentation to extract all keywords of each financing product to obtain a second keyword, the first keyword is converted into a word vector to obtain a first word vector, the second keyword is converted into a word vector to obtain a second word vector, a cosine distance between the first word vector and the second word vector is calculated, the cosine distance is used as a similarity value between the first keyword and the second keyword of each financing product, and then the second keyword with a similarity value greater than a preset similarity threshold value is used as a target keyword, and taking the financing product corresponding to the target keyword as a target financing product, thereby screening the target financing product which accords with the current affair hotspot, recommending the target financing product to the user and improving the timeliness of recommendation.
In an embodiment, the calculating the similarity value between the first keyword and the second keyword of each financial product in step S162 may specifically include:
s1621, respectively converting the first keywords into Word vectors by using a Word2Vec Word vector model trained in advance, and calculating an average value of the Word vectors of the first keywords to obtain a first Word vector;
s1622, converting the second keyword of each financial product into a Word vector by using the Word2Vec Word vector model to obtain a second Word vector of each financial product;
s1623, calculating the cosine distance between the first word vector and the second word vector of each financial product to serve as the similarity value.
In this embodiment, a Word2Vec Word vector model trained in advance is used to convert the plurality of first keywords into Word vectors respectively, so as to obtain Word vectors corresponding to the first keywords, and an average value of the Word vectors of all the first keywords is used as the first Word vector; and meanwhile, converting the second keywords into word vectors to obtain second word vectors, calculating cosine distances between the first word vectors and the second word vectors, and taking the cosine distances as similarity values of the first keywords and the second keywords of each financial product so as to accurately calculate the similarity values of the first keywords and the second keywords of each financial product.
The Word2Vec Word vector model is a model for learning semantic knowledge from a large amount of texts and adopts an unsupervised mode. The method is characterized in that a large amount of texts are trained, words in the texts are represented in a vector form, the vector is called a word vector, and the relation between two words can be known by calculating the distance between the word vectors of the two words.
In an embodiment, the obtaining of the trained K decision tree models may specifically include:
respectively calculating the loss value of each trained decision tree model by using a preset loss function;
judging whether the loss value of each decision tree model is lower than a preset loss value or not;
and if so, judging that the K decision tree models finish training.
In this embodiment, after each training of the decision tree model, a preset loss function may be used to calculate a loss value of the decision tree model after each training is completed, and when the loss value meets a preset threshold or is smaller than the preset loss value, that is, meets a requirement, it is indicated that the decision tree model meets the training requirement, and the training of the decision tree model is completed, so as to improve the recommendation accuracy of the financial product recommendation model.
When the loss value of any decision tree model is not less than the preset loss value, forward transmission can be performed in a neural network structure of the decision tree model according to the loss value, relevant parameters of the decision tree model are adjusted, the adjusted decision tree model is retrained based on the reset relevant parameters, retraining is calculated until the loss values of all the decision tree models are less than the preset loss value, and when the loss value meets the preset requirement, parameters of the decision tree model corresponding to the loss value meeting the preset threshold value are finally obtained, so that training of all the decision tree models is finished.
In an embodiment, the screening out the target financing product from the financing products according to the product preference value may specifically include:
sorting the financial products in the order of the secondary magnitude of the product preference values;
taking the front N position of the financing products as target financing products; wherein N is a positive integer.
In this embodiment, the financial products are sorted based on the product preference values output by the financial product recommendation model, and when sorting is performed, all the financial products are sequentially sorted according to the descending order of the product preference values, and then the top N financial products are screened out from all the financial products as target financial products, so as to screen out the financial products most suitable for users, for example, the top three financial products with product preference values are screened out as target financial products. The product preference value is used for predicting the preference information degree of the user on the output financing product, and the larger the product preference value is, the larger the probability that the user likes or fits the financing product is.
In an embodiment, the acquiring historical financial data of the user within the preset time period includes:
capturing a financing page browsed by the user within a preset time period through a crawling tool;
extracting information in the financing page by adopting a preset regular expression to obtain target information;
and analyzing the target information to obtain the historical financing data.
The financing page refers to a webpage related to a financing product, and the financing page in the embodiment may be a webpage of a finance website or a webpage of a stock website. Specifically, the website of the financing page can be read through a read () method, a website is transferred to a getHtml () function, and the whole page is downloaded to obtain the financing page.
The preset regular expression is a character string matching and processing rule and is used for extracting information in the webpage. The preset regular expression includes, but is not limited to, a Python regular expression. The target information refers to webpage information matched with the regular expression. The target information may be information such as the name of the financial product, the income, or the product type.
Specifically, information matched with a preset regular expression is filtered from the financial page, and then the information is extracted to obtain target information, so that the information in the financial page is extracted by the preset regular expression, and the accuracy of the target information is improved.
In one embodiment, when analyzing the target information, firstly, data analysis is performed on the target information through an analysis module in a crawler analysis library, then the analyzed target information is extracted through a path expression, and the analyzed target information is stored in a database to obtain historical financial data. The crawler analysis library can be a Beautiful Soup analysis library or an lxml analysis library. Understandably, historical financial data can be quickly and accurately acquired by analyzing the target information.
According to the embodiment, the information in the financing webpage is extracted by adopting the preset regular expression, so that the accuracy of the target information is improved; and finally, analyzing the target information to obtain historical financial data, thereby quickly and accurately obtaining the historical financial data.
In addition, when a user logs in a system page, the user is prompted to register description information such as identity, mobile phone number and the like; the identity information of the user can be stored in a background oracle database so as to be convenient for use of information security, subsequent recommendation information and the like. The system background oracle database records the browsing product id duration of a user, the concerned product id, the purchased product id and the removed product id to generate financial product data, and the financial product data can be subsequently taken for training, so that the daily financial product data of the oracle are synchronized to a big data platform and are stored according to daily partitions.
In one embodiment, the obtaining the historical financial data includes:
dividing the analyzed target information into a plurality of sub-data blocks, and respectively allocating a unique identifier for each sub-data block according to a data structure;
and traversing all the identifiers, reserving a unique data block corresponding to each identifier, removing other data blocks with the same identifier, and obtaining the historical financial data with the repeated data removed.
In this embodiment, the parsed target information is divided into a plurality of sub-data blocks, each of which has the same size, and a unique identifier is respectively allocated to each sub-data block according to the data structure, where the identifier is used to distinguish different data blocks, so that the same data block has the same identifier, and then the identifier is used to remove duplicate data, so as to obtain historical financial data from which the duplicate data is removed. Specifically, all identifiers may be traversed, a unique data block is reserved corresponding to each identifier, and other data blocks having the same identifier are removed.
The data structure refers to a collection of data elements having one or more relationships with each other and relationship components among the data elements in the collection. Common data structures are: arrays, stacks, linked lists, queues, trees, graphs, heaps, hash tables, and the like.
Referring to fig. 2, an embodiment of the present application further provides a financial product recommendation device based on artificial intelligence, including:
the acquisition module 11 is used for acquiring historical financial data and historical product preference information of a user in a preset time period and establishing a corresponding relation between the historical financial data and the historical product preference information;
the dividing module 12 is configured to divide historical financing data with the same historical product preference information into the same data set according to the corresponding relationship, and form a training set by using all data sets and historical product preference information corresponding to the data sets as training samples;
the training module 13 is configured to select K training samples from the training set, construct K decision tree models, and input each training sample to a different decision tree model for training, so as to obtain K trained decision tree models, where K is a positive integer;
the composition module 14 is used for forming a financial product recommendation model by using the trained K decision tree models;
the input module 15 is used for acquiring financial data of a user, inputting the financial data into the financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product;
and the recommending module 16 is used for screening out a target financing product from the financing products according to the product preference value and recommending the target financing product to a user.
Historical financial data may include age, income, marital status, academic history, asset level, financial product type, risk level, investment period, amount of money invested, profitability, and the like. The historical product preference information includes behavioral data such as preferred investment funds, stocks or bonds, and the like.
The corresponding relation between the historical financing data and the historical product preference information is constructed for better dividing the historical financing data, for example, when the financing product type is that the occupation ratio of bonds is maximum in the financing products of the user based on the historical financing data analysis, the historical product preference information of the user is a favorite investment bond and belongs to a robust type, and the corresponding historical financing data is data related to the bond, including the investment amount, holding time and the like of the bond.
The embodiment classifies historical financing data based on the corresponding relation, divides the historical financing data with the same historical product preference information into the same data set, and takes all the data sets and the historical product preference information corresponding to the data sets as training samples to form a training set. For example, all data about bond investment belonging to favorite investment bonds are divided into a data set, the same data set and corresponding historical product preference information are used as a training sample, and a plurality of training samples are combined into a training set and stored.
The embodiment can use a machine learning model algorithm to recommend and match financial products, and meets the core business target. Before recommendation, a financial product recommendation model needs to be obtained through training. Specifically, the decision tree model can be used for carrying out classification learning on the training set, the rule that real customers buy the financial products and the corresponding product preference values are mined, the mapping relation between the real customers and the product preference values is established, and finally the corresponding financial product recommendation model is generated. For example, K training samples can be randomly extracted from a training set, K decision tree models are also constructed, each decision tree model inputs a unique training sample, the data quantity of historical financing data of each training sample is enough, multiple times of training are conducted on the corresponding decision tree models to obtain the trained decision tree models, each trained decision tree model can correspondingly identify a product preference information based on the financing data, a financing product recommendation model with a better recommendation effect is obtained through a separate training mode, training in the same time period can be achieved through an independent training mode, and training efficiency is improved.
In addition, the K decision tree classification models can be randomly combined to obtain a trained financial product recommendation model, for example, the K decision tree classification models can be spliced to be combined into the trained financial product recommendation model, or at least two of the K decision tree classification models can be selected to be spliced to be combined into the trained financial product recommendation model, so that the product recommendation model obtained by combining the separately trained decision tree classification models has a better training result, and the recommendation precision of financial products is improved.
The financial product recommendation model can also calculate the product preference value of each financial product based on historical product preference information, and the product preference value is used for evaluating the preference degree of a user for the financial products. Therefore, the financial product recommendation model can output the product preference value corresponding to each financial product besides the financial product needing to be recommended. Specifically, the financial data of a user needing to recommend a financial product is obtained, the financial data is input into a trained financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product, then a target financial product is screened out from the financial products according to the product preference values, and the target financial product is recommended to the user. For example, a financing product with the highest product preference value is screened out as a target financing product, and the target financing product is recommended to the user.
As described above, it can be understood that each component of the financial product recommendation device based on artificial intelligence proposed in the present application can implement the function of any one of the above financial product recommendation methods based on artificial intelligence, and the detailed structure is not repeated.
Referring to fig. 3, an embodiment of the present application further provides a computer device, and an internal structure of the computer device may be as shown in fig. 3. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the computer designed processor is used to provide computational and control capabilities. The memory of the computer device comprises a storage medium and an internal memory. The storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and computer programs in the storage medium. The database of the computer equipment is used for storing historical financial data, historical product preference information and other data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement an artificial intelligence based financial product recommendation method.
The processor executes the method for recommending the financial product based on the artificial intelligence, and comprises the following steps:
acquiring historical financing data and historical product preference information of a user in a preset time period, and constructing a corresponding relation between the historical financing data and the historical product preference information;
according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, and taking all data sets and the historical product preference information corresponding to the data sets as training samples to form a training set; selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
forming a financial product recommendation model by using the trained K decision tree models;
acquiring financial data of a user, and inputting the financial data into the financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product;
and screening out a target financing product from the financing products according to the product preference value, and recommending the target financing product to a user.
An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, the computer program, when executed by a processor, implementing a financial product recommendation method based on artificial intelligence, including the steps of:
acquiring historical financing data and historical product preference information of a user in a preset time period, and constructing a corresponding relation between the historical financing data and the historical product preference information;
according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, and taking all data sets and the historical product preference information corresponding to the data sets as training samples to form a training set; selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
forming a financial product recommendation model by using the trained K decision tree models;
acquiring financial data of a user, and inputting the financial data into the financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product;
and screening out a target financing product from the financing products according to the product preference value, and recommending the target financing product to a user.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and can include the processes of the embodiments of the methods described above when the computer program is executed. Any reference to memory, storage, database, or other medium provided herein and used in the examples may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), double-rate SDRAM (SSRSDRAM), Enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and bus dynamic RAM (RDRAM).
To sum up, the most beneficial effect of this application lies in:
according to the financial product recommendation method, device, equipment and medium based on artificial intelligence, the corresponding relation between historical financial data and historical product preference information is established by acquiring the historical financial data and the historical product preference information of a user in a preset time period; according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, taking all data sets and historical product preference information corresponding to the data sets as training samples to form a training set, selecting K training samples from the training set, constructing K decision tree models, and inputting each training sample to different decision tree models for training to obtain K trained decision tree models; forming a financial product recommendation model by using the trained K decision tree models; acquiring financial data of a user, inputting the financial data into a financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product; and screening out target financing products from the financing products according to the product preference values, and recommending the target financing products to the user. The method comprises the steps of classifying historical financing data based on reflected historical product preference information, correspondingly training a decision tree model by utilizing each type of historical financing data, realizing independent training, improving training efficiency, and obtaining a financing product recommendation model by utilizing a combination of individually trained decision tree classification models, so that the generated financing product recommendation model has a better training result, and the recommendation precision of the financing product is improved; meanwhile, the financial products are screened based on the product preference values output by the financial product recommendation model, so that financial products suitable for users are screened out, accurate recommendation is achieved, and the recommendation effect is better.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, apparatus, article, or method that includes the element.
The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are also included in the scope of the present application.
Claims (10)
1. A financial product recommendation method based on artificial intelligence is characterized by comprising the following steps:
acquiring historical financing data and historical product preference information of a user in a preset time period, and constructing a corresponding relation between the historical financing data and the historical product preference information;
according to the corresponding relation, dividing historical financing data with the same historical product preference information into the same data set, and taking all data sets and the historical product preference information corresponding to the data sets as training samples to form a training set;
selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
forming a financial product recommendation model by using the trained K decision tree models;
acquiring financial data of a user, and inputting the financial data into the financial product recommendation model to obtain a plurality of financial products and product preference values corresponding to each financial product;
and screening out a target financing product from the financing products according to the product preference value, and recommending the target financing product to a user.
2. The method of claim 1, wherein said screening out a target financial product from said financial products based on said product preference value comprises:
acquiring current affair hotspot information, extracting a first keyword of the current affair hotspot information, and extracting a second keyword of each financial product;
calculating the similarity value of the first keyword and a second keyword of each financial product, and taking the second keyword with the similarity value larger than a preset similarity threshold value as a target keyword;
and taking the financing product corresponding to the target keyword as the target financing product.
3. The method of claim 2, wherein said first keyword comprises a plurality of keywords, and wherein said calculating a similarity value between said first keyword and a second keyword of each of said financial products comprises:
respectively converting the first keywords into Word vectors by using a Word2Vec Word vector model trained in advance, and calculating the average value of the Word vectors of the first keywords to obtain a first Word vector;
converting the second keyword of each financial product into a Word vector by using the Word2Vec Word vector model to obtain a second Word vector of each financial product;
and calculating the cosine distance between the first word vector and the second word vector of each financial product as the similarity value.
4. The method of claim 1, wherein the obtaining of the trained K decision tree models comprises:
respectively calculating the loss value of each trained decision tree model by using a preset loss function;
judging whether the loss value of each decision tree model is lower than a preset loss value or not;
and if so, judging that the K decision tree models finish training.
5. The method of claim 1, wherein said screening out a target financial product from said financial products based on said product preference value comprises:
sorting the financial products in the order of the secondary magnitude of the product preference values;
taking the front N position of the financing products as target financing products; wherein N is a positive integer.
6. The method of claim 1, wherein the obtaining historical financial data of the user within the preset time period comprises:
capturing a financing page browsed by the user within a preset time period through a crawling tool;
extracting information in the financing page by adopting a preset regular expression to obtain target information;
and analyzing the target information to obtain the historical financing data.
7. The method of claim 6, wherein said obtaining said historical financial data comprises:
dividing the analyzed target information into a plurality of sub-data blocks, and respectively allocating a unique identifier for each sub-data block according to a data structure;
and traversing all the identifiers, reserving a unique data block corresponding to each identifier, removing other data blocks with the same identifier, and obtaining the historical financial data with the repeated data removed.
8. A financial product recommendation device based on artificial intelligence, comprising:
the acquisition module is used for acquiring historical financing data and historical product preference information of a user in a preset time period and constructing a corresponding relation between the historical financing data and the historical product preference information;
the dividing module is used for dividing historical financing data with the same historical product preference information into the same data set according to the corresponding relation, and forming a training set by taking all data sets and the historical product preference information corresponding to the data sets as training samples; the training module is used for selecting K training samples from the training set, constructing K decision tree models, and inputting the training samples into different decision tree models for training to obtain K trained decision tree models, wherein K is a positive integer;
the composition module is used for forming a financial product recommendation model by utilizing the trained K decision tree models;
the input module is used for acquiring financial data of a user, inputting the financial data into the financial product recommendation model, and obtaining a plurality of financial products and product preference values corresponding to each financial product;
and the recommending module is used for screening out a target financing product from the financing products according to the product preference value and recommending the target financing product to a user.
9. A computer device, comprising:
a processor;
a memory;
a computer program, wherein the computer program is stored in the memory and configured to be executed by the processor, the computer program configured to perform the artificial intelligence based financial product recommendation method of any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, implements the artificial intelligence based financial product recommendation method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111052153.7A CN113762392A (en) | 2021-09-08 | 2021-09-08 | Financial product recommendation method, device, equipment and medium based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111052153.7A CN113762392A (en) | 2021-09-08 | 2021-09-08 | Financial product recommendation method, device, equipment and medium based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113762392A true CN113762392A (en) | 2021-12-07 |
Family
ID=78793981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111052153.7A Pending CN113762392A (en) | 2021-09-08 | 2021-09-08 | Financial product recommendation method, device, equipment and medium based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113762392A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114297511A (en) * | 2022-01-27 | 2022-04-08 | 中国农业银行股份有限公司 | Financing recommendation method, device, system and storage medium |
| CN115511569A (en) * | 2022-10-09 | 2022-12-23 | 湖南工商大学 | Financial risk monitoring device and method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130046772A1 (en) * | 2011-08-16 | 2013-02-21 | Alibaba Group Holding Limited | Recommending content information based on user behavior |
| CN105631707A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Advertisement click rate estimation method based on decision tree, application recommendation method and device |
| CN108629693A (en) * | 2018-05-08 | 2018-10-09 | 平安科技（深圳）有限公司 | Automatically generate method, apparatus, computer equipment and the storage medium of suggestion for investment |
| CN110752942A (en) * | 2019-09-06 | 2020-02-04 | 平安科技（深圳）有限公司 | Alarm information decision method and device, computer equipment and storage medium |
| CN111861569A (en) * | 2020-07-23 | 2020-10-30 | 中国工商银行股份有限公司 | Product information recommendation method and device |
| US20210118080A1 (en) * | 2018-04-18 | 2021-04-22 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for recommending transportation means |
| CN113255370A (en) * | 2021-06-22 | 2021-08-13 | 中国平安财产保险股份有限公司 | Industry type recommendation method, device, equipment and medium based on semantic similarity |
-
2021
- 2021-09-08 CN CN202111052153.7A patent/CN113762392A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130046772A1 (en) * | 2011-08-16 | 2013-02-21 | Alibaba Group Holding Limited | Recommending content information based on user behavior |
| CN105631707A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Advertisement click rate estimation method based on decision tree, application recommendation method and device |
| US20210118080A1 (en) * | 2018-04-18 | 2021-04-22 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for recommending transportation means |
| CN108629693A (en) * | 2018-05-08 | 2018-10-09 | 平安科技（深圳）有限公司 | Automatically generate method, apparatus, computer equipment and the storage medium of suggestion for investment |
| CN110752942A (en) * | 2019-09-06 | 2020-02-04 | 平安科技（深圳）有限公司 | Alarm information decision method and device, computer equipment and storage medium |
| CN111861569A (en) * | 2020-07-23 | 2020-10-30 | 中国工商银行股份有限公司 | Product information recommendation method and device |
| CN113255370A (en) * | 2021-06-22 | 2021-08-13 | 中国平安财产保险股份有限公司 | Industry type recommendation method, device, equipment and medium based on semantic similarity |
Non-Patent Citations (1)
| Title |
|---|
| 辛菊琴;蒋艳;舒少龙;: "综合用户偏好模型和BP神经网络的个性化推荐", 计算机工程与应用, no. 02, pages 57 - 60 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114297511A (en) * | 2022-01-27 | 2022-04-08 | 中国农业银行股份有限公司 | Financing recommendation method, device, system and storage medium |
| CN115511569A (en) * | 2022-10-09 | 2022-12-23 | 湖南工商大学 | Financial risk monitoring device and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110765265B (en) | Information classification extraction method and device, computer equipment and storage medium | |
| CN110175227B (en) | Dialogue auxiliary system based on team learning and hierarchical reasoning | |
| CN110147445A (en) | Intension recognizing method, device, equipment and storage medium based on text classification | |
| CN111221944B (en) | Text intention recognition method, device, equipment and storage medium | |
| CN110458324B (en) | Method and device for calculating risk probability and computer equipment | |
| CN107943792B (en) | Statement analysis method and device, terminal device and storage medium | |
| CN111274371B (en) | Intelligent man-machine conversation method and equipment based on knowledge graph | |
| CN111090719B (en) | Text classification method, apparatus, computer device and storage medium | |
| De Souza et al. | Automated multi-label text categorization with VG-RAM weightless neural networks | |
| CN112395500A (en) | Content data recommendation method and device, computer equipment and storage medium | |
| CN113762392A (en) | Financial product recommendation method, device, equipment and medium based on artificial intelligence | |
| CN112084307A (en) | Data processing method and device, server and computer readable storage medium | |
| CN114647713A (en) | Knowledge graph question-answering method, device and storage medium based on virtual confrontation | |
| CN114510923B (en) | Text theme generation method, device, equipment and medium based on artificial intelligence | |
| CN111460114A (en) | Retrieval method, device, equipment and computer readable storage medium | |
| CN114936326A (en) | Information recommendation method, device, equipment and storage medium based on artificial intelligence | |
| CN114238798A (en) | Search ranking method, system, device and storage medium based on neural network | |
| CN114707510A (en) | Resource recommendation information pushing method and device, computer equipment and storage medium | |
| CN111400413B (en) | Method and system for determining category of knowledge points in knowledge base | |
| CN116541517A (en) | Text information processing method, apparatus, device, software program, and storage medium | |
| CN114283429A (en) | Material work order data processing method, device, equipment and storage medium | |
| CN113869068A (en) | Scene service recommendation method, device, equipment and storage medium | |
| CN113688636A (en) | Extended question recommendation method and device, computer equipment and storage medium | |
| CN111611379A (en) | Text information classification method, device, equipment and readable storage medium | |
| CN112148855A (en) | Intelligent customer service problem retrieval method, terminal 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 |