CN111833113A - Product recommendation method, device and equipment based on big data and storage medium - Google Patents

Abstract

The invention relates to big data and discloses a product recommendation method, device and equipment based on the big data and a storage medium. The big data based product recommendation method comprises the following steps: reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes; inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products; screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme; and packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account. In addition, the invention also relates to a block chain technology, and the user account information can be stored in the block chain. Through the analysis of each behavior index of the user, the accuracy of generating the product recommendation scheme for the user is higher.

Description

Product recommendation method, device and equipment based on big data and storage medium

Technical Field

The invention relates to big data intelligent recommendation, in particular to a product recommendation method, device, equipment and storage medium based on big data.

Background

Since the eighties of the last century, the service industry has been developed at a high speed, and with the development of the internet industry, the contact area of people to various service industries, such as financial products, insurance industries and the like, is wider and wider, the awareness of risk guarantee is also improved, and the types of service products are increasingly diversified. With the rapid development of internet finance, the service type product types are increased in an explosive manner, the recommended advertisements of the service type products are visible everywhere, the product APP is opened, various service type products can be seen, and the transaction function can be realized by one key.

Software is one of the most direct recommending ways for service type products, the existing service type product recommending method mainly selects products with better sales quantity to be placed on the home page of the product according to historical sales records or recommends according to purchase records of users, but the asset condition, the purchase demand and the like of each customer are different, and the mode cannot really meet all consumers. For a company, whether according to historical sales records or according to purchasing habits of users, relevant service type products are placed in a service type product matching scheme for consumers and recommended to the consumers, obvious consideration factors are not comprehensive enough, in other words, accuracy is not enough.

Disclosure of Invention

The invention mainly aims to solve the technical problem that the recommendation accuracy of the service type product is not enough at present.

The invention provides a product recommendation method based on big data in a first aspect, which comprises the following steps:

reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

and packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

Optionally, in a first implementation manner of the first aspect of the present invention, before the inputting the user account information into a preset preference index evaluation model, and counting scores of the multiple user behavior indexes through the preference index evaluation model to obtain a first preference index corresponding to satisfaction of the user account with various types of products, the method further includes:

acquiring account information of a whole network user and storing the account information in a message queue in a distributed manner;

every other preset period, acquiring a corresponding user behavior index from the message queue, and counting the distribution of the user behavior index in the whole network;

calculating the distribution interval of the scores of the user behavior indexes based on the distribution of the user behavior indexes in the whole network to obtain a calculation result;

and determining a preference index evaluation rule of the user account for various products based on the calculation result, wherein the preference index evaluation rule is applied to the preference index evaluation model for counting the scores of the user behavior indexes.

Optionally, in a second implementation manner of the first aspect of the present invention, the preference index evaluation model includes a plurality of evaluation indexes, where the evaluation indexes include an account opening model factor, a transaction model factor, a browsing model factor, a fund model factor, an evaluation model factor, a logout model factor, an encryption model factor, a credit model factor, and a wind control model factor;

wherein the preference index evaluation rule comprises:

determining the score of the account opening model factor according to the account opening time, wherein the longer the account opening time is, the lower the score of the account opening model factor is; or

Determining the grade of the transaction model factor according to the transaction time and the transaction amount of the account, wherein the closer the transaction time is, the larger the transaction amount is, and the higher the grade of the transaction model factor is; or

Determining the score of the browsing model factor according to the time length and the number of the products browsed by the account, wherein the score of the browsing model factor is higher when the time length of browsing the products is larger and the number of the products is larger; or

Determining the score of the fund model factor according to the account fund remaining quantity and the remaining time, wherein the larger the remaining quantity is, the longer the remaining time is, the higher the score of the fund model factor is; or

If the user account is evaluated, the evaluation model factor is fully scored, otherwise, the score is not added; or

If the account has no log-off record, logging off the model factor to obtain full score, otherwise, not adding score; or

If the user account is encrypted by biological identification, the encryption model factor is fully scored, otherwise, the score is not added; or

If the account is not on the blacklist, the credit model factor is fully scored, otherwise, the score is not added; or

And determining the score of the wind control model factor according to the wind control level of the account, wherein the less the wind control level is, the higher the score of the wind control model factor is.

Optionally, in a third implementation manner of the first aspect of the present invention, the screening, based on the first preference index, products corresponding to the user account to obtain a corresponding product recommendation scheme includes:

acquiring a product transaction record, wherein the transaction record comprises transaction product information and transaction account information;

inputting the preference index evaluation model according to the transaction account information, and counting the scores of the user behavior indexes according to the preference index evaluation model to obtain a corresponding second preference index;

calculating the difference value between the first preference index and the second preference index, and screening a preset number of corresponding transaction accounts in a descending order according to the difference value;

and screening the transaction product information corresponding to the transaction account to obtain a corresponding product recommendation scheme.

Optionally, in a fourth implementation manner of the first aspect of the present invention, after the packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account, the method further includes:

monitoring whether a client corresponding to the user account uploads feedback information corresponding to the product recommendation scheme;

if yes, obtaining the feedback information, and extracting characteristic information in the feedback information;

determining a correction value corresponding to the product recommendation scheme based on the characteristic information;

and adjusting the product recommendation scheme corresponding to the user account based on the correction value.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the product recommendation method further includes:

determining a family member structure corresponding to the user account according to the user account information;

judging whether a family associated account corresponding to the family member structure exists or not;

if so, acquiring a product recommendation scheme corresponding to the family associated account;

and screening a preset number of products according to the product recommendation sequence of the product recommendation scheme corresponding to the family associated account, and randomly adding the products into the product recommendation scheme corresponding to the user account.

Optionally, in a sixth implementation manner of the first aspect of the present invention, after the screening, based on the first preference index, products corresponding to the user account to obtain a corresponding product recommendation scheme, the method further includes:

counting the grade of each evaluation index in the user account, and selecting the evaluation index with the highest grade as the advantage evaluation index of the user account;

screening a preset number of products as dominant recommended products of the user account according to the product recommendation sequence of the product recommendation scheme;

and determining a corresponding product label according to the advantage evaluation index and the advantage recommended product so as to identify the corresponding advantage recommended product when the product recommendation scheme is pushed.

The invention provides a big data-based product recommendation device in a second aspect, which comprises:

the system comprises a reading module, a recommendation module and a recommendation module, wherein the reading module is used for reading user account information of a product to be recommended, and the user account information comprises a plurality of user behavior indexes;

the index evaluation module is used for inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

the scheme generation module is used for screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

and the scheme pushing module is used for packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

Optionally, in a first implementation manner of the second aspect of the present invention, the big-data-based product recommendation apparatus further includes:

the acquisition module is used for acquiring the account information of the users in the whole network and storing the account information in a message queue in a distributed manner;

the statistical module is used for acquiring corresponding user behavior indexes from the message queue every other preset period and counting the distribution of the user behavior indexes in the whole network;

the calculation module is used for calculating the distribution interval of the scores of the user behavior indexes based on the distribution of the user behavior indexes in the whole network to obtain a calculation result;

and the evaluation rule generating module is used for determining a preference index evaluation rule of the user account for various products based on the calculation result, wherein the preference index evaluation rule is applied to the preference index evaluation model for counting the scores of the user behavior indexes.

Optionally, in a second implementation manner of the second aspect of the present invention, the preference index evaluation model includes a plurality of evaluation indexes, where the evaluation indexes include an account opening model factor, a transaction model factor, a browsing model factor, a fund model factor, an evaluation model factor, a logout model factor, an encryption model factor, a credit model factor, and a wind control model factor;

wherein the preference index evaluation rule comprises:

determining the score of the account opening model factor according to the account opening time, wherein the longer the account opening time is, the lower the score of the account opening model factor is; or

Determining the grade of the transaction model factor according to the transaction time and the transaction amount of the account, wherein the closer the transaction time is, the larger the transaction amount is, and the higher the grade of the transaction model factor is; or

Determining the score of the browsing model factor according to the time length and the number of the products browsed by the account, wherein the score of the browsing model factor is higher when the time length of browsing the products is larger and the number of the products is larger; or

Determining the score of the fund model factor according to the account fund remaining quantity and the remaining time, wherein the larger the remaining quantity is, the longer the remaining time is, the higher the score of the fund model factor is; or

If the user account is evaluated, the evaluation model factor is fully scored, otherwise, the score is not added; or

If the account has no log-off record, logging off the model factor to obtain full score, otherwise, not adding score; or

If the user account is encrypted by biological identification, the encryption model factor is fully scored, otherwise, the score is not added; or

If the account is not on the blacklist, the credit model factor is fully scored, otherwise, the score is not added; or

And determining the score of the wind control model factor according to the wind control level of the account, wherein the less the wind control level is, the higher the score of the wind control model factor is.

Optionally, in a third implementation manner of the second aspect of the present invention, the scheme generating module further includes:

the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring a product transaction record, and the transaction record comprises transaction product information and transaction account information;

the statistic unit is used for inputting the preference index evaluation model according to the transaction account information, and carrying out statistics on scores of the user behavior indexes according to the preference index evaluation model to obtain a corresponding second preference index;

the calculating unit is used for calculating the difference value between the first preference index and the second preference index and screening a preset number of corresponding transaction accounts in a descending order according to the difference value;

and the generating unit is used for screening the transaction product information corresponding to the transaction account to obtain a corresponding product recommendation scheme.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the big-data-based product recommendation apparatus further includes:

the monitoring module is used for monitoring whether the client corresponding to the user account uploads the feedback information corresponding to the product recommendation scheme;

the extraction module is used for acquiring the feedback information and extracting the characteristic information in the feedback information if the feedback information corresponding to the product recommendation scheme is uploaded by the client corresponding to the user account;

the correction module is used for determining a correction value corresponding to the product recommendation scheme based on the characteristic information; and adjusting the product recommendation scheme corresponding to the user account based on the correction value.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the product recommendation apparatus further includes an association module, where the association module is specifically configured to:

determining a family member structure corresponding to the user account according to the user account information;

judging whether a family associated account corresponding to the family member structure exists or not;

if so, acquiring a product recommendation scheme corresponding to the family associated account;

and screening a preset number of products according to the product recommendation sequence of the product recommendation scheme corresponding to the family associated account, and randomly adding the products into the product recommendation scheme corresponding to the user account.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the product recommendation apparatus further includes an identification module, where the identification module is specifically configured to:

counting the grade of each evaluation index in the user account, and selecting the evaluation index with the highest grade as the advantage evaluation index of the user account;

screening a preset number of products as dominant recommended products of the user account according to the product recommendation sequence of the product recommendation scheme;

and determining a corresponding product label according to the advantage evaluation index and the advantage recommended product so as to identify the corresponding advantage recommended product when the product recommendation scheme is pushed.

The invention provides a product recommendation device based on big data in a third aspect, which comprises: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line; the at least one processor invokes the instructions in the memory to cause the big-data based product recommendation device to perform the big-data based product recommendation method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute the big data based product recommendation method described above.

In the technical scheme of the invention, user account information of a product to be recommended is read, wherein the user account information comprises a plurality of user behavior indexes; inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products; screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme; and packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account. By analyzing the historical behavior index record information of the user in the user account, a product recommendation scheme more suitable for the corresponding user is generated, and the recommendation result accuracy is higher.

Drawings

FIG. 1 is a schematic diagram of a big data-based product recommendation method according to a first embodiment of the present invention;

FIG. 2 is a schematic diagram of a second embodiment of the big data based product recommendation method of the present invention;

FIG. 3 is a schematic diagram of a third embodiment of the big data based product recommendation method according to the present invention;

FIG. 4 is a schematic diagram of a fourth embodiment of the big data based product recommendation method according to the present invention;

FIG. 5 is a schematic diagram of an embodiment of a big data based product recommendation device of the present invention;

FIG. 6 is a schematic diagram of another embodiment of a big data based product recommendation device of the present invention;

FIG. 7 is a schematic diagram of an embodiment of a big data based product recommendation device according to the present invention.

Detailed Description

The technical scheme includes that a plurality of behavior indexes of a user account are used as evaluation parameters of a product recommendation scheme, the parameters of the behavior indexes are counted firstly to obtain product purchase preference of the account, a preset number of adaptive product combinations are screened according to the quantized account purchase preference to serve as the product recommendation scheme of the account, the product recommendation scheme is packaged to be displayed when the account logs in, the behavior indexes of the user are comprehensively evaluated, and besides the purchasing power of the user, the product recommendation method, the device and the equipment also have personal quality analysis to provide the most appropriate product recommendation scheme to the corresponding user.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a detailed flow of an embodiment of the present invention is described below, and referring to fig. 1, a first embodiment of a big data based product recommendation method in an embodiment of the present invention includes:

101. reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

it is to be understood that the executing subject of the present invention may be a big data based product recommendation device, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject. The product type intelligently recommended by the invention is a service type product, mainly comprises insurance products, financial products, securities, funds, stocks and the like, and preferably, the specific implementation process of the invention is explained by recommending the insurance products. It is emphasized that, in order to further ensure the privacy and security of the user account information, the user account information may also be stored in a node of a block chain.

In this embodiment, the user includes, in the software-like user account information, user behavior indexes such as account opening time, transaction record, browsing record, fund retention record, wind control level, whether the user has ever logged off the account, whether a biometric password is set, whether evaluation is completed, whether the user behavior indexes are in a blacklist, and the like, that is, the user uses the account to perform a consumption behavior since registering the account.

The data storage format of each user behavior index is as follows:

opening time: "year, month, day/hour, minute, second", such as 20100115/161359;

and (3) transaction recording: each transaction record is rewound and arranged from far to near according to time, and the data storage format is 'year, month, day/hour, minute and second, transaction amount', such as 20200115/161359, 250;

and (4) browsing records: each insurance product browsing record is rewound and is arranged from far to near according to time, and the data storage format is 'browsing duration and browsing times', such as '1 h35min8s, 9';

and (4) fund retention record: "retained balance, retained time (not less than the retained balance counted in days)", such as "25000, 78 d";

wind control grade: "wind control level", such as "level three";

whether the evaluation is finished: whether the mark is 1 or 0;

whether or not the account was logged off: whether the mark is 1 or 0;

whether a biological password is set: whether the mark is 1 or 0;

whether it is in the blacklist: whether the mark is 1 or 0;

102. inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

in this embodiment, the preset preference index evaluation rule is used as a score reference for each user behavior index, and a user behavior index interval corresponding to each score is defined, wherein after the user account information is input into the preset preference index model, one or more rules in the preference index evaluation rule can be used for calculation. And calculating the score of each user behavior index in the user account information according to the threshold preference index evaluation rule, accumulating the scores, and taking the accumulated result as a first preference index of the user on various products, wherein the first preference index represents the consumption capability, consumption habit and account current situation of the user so as to predict the product type of the user possibly mental apparatus.

In this embodiment, by presetting the preference index evaluation rule, the score of each user behavior index can be obtained as follows:

opening time: a;

and (3) transaction recording: b;

and (4) browsing records: c;

and (4) fund retention record: d;

wind control grade: e;

whether the evaluation is finished: f;

whether or not the account was logged off: g;

whether a biological password is set: h;

whether it is in the blacklist: i;

the corresponding first preference index dex1 ═ a + b + c + d + e + f + g + h + i.

103. Screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

in this embodiment, the first preference index reflects the consumption ability, consumption habits and consumption status of the user, and the insurance products can be related to each other according to evaluation factors such as risk level, income recovery mode and storage duration, and the types of the insurance products suitable for the user can be evaluated to a certain extent through the first preference index, for example, an account with a high first preference index can be recommended to the user with a high risk level, high income, long-term investment types and a long storage duration; and users with a low first preference index can recommend insurance products with low risk level, robust income, short-term investment and short survival time for the users.

In the screening process, the first preference index is used as a median, products suitable for the user are screened within a preset interval range, for example, if the first preference index dex1 of the user is equal to a, the corresponding products are screened through a first preference index interval (a-b, a + b), and if a is less than b, the first preference index interval is (0, a + b).

104. And packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

In this embodiment, according to the first preference index obtained by account information statistics, the product recommendation scheme screened out to be suitable for the user is not pushed to the user in time, but is packaged into a file package, and when the user logs in and browses insurance software, recommendation is displayed on the home page. Here the package file is associated with the user account display front end through an interface.

In this embodiment, when the user logs in the account in the insurance software, the user initiates a login request to the server, the server determines the product recommendation scheme file package corresponding to the user through the user field, the interface authority is used as a limit, and then data is returned to the front end according to the user interface authority so that the front end can render the page and generate a paired user account home page.

In the embodiment of the invention, a plurality of behavior indexes of a user account are used as evaluation parameters of a product recommendation scheme, the parameters of the behavior indexes are counted to obtain the product purchase preference of the account, a preset number of matched products are screened according to the quantified account purchase preference to combine the product recommendation scheme of the account, and then the product recommendation scheme is packaged for displaying the product recommendation scheme when the account logs in.

Referring to fig. 2, a second embodiment of the big data based product recommendation method according to the embodiment of the present invention includes:

201. reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

202. acquiring account information of a whole network user and storing the account information in a message queue in a distributed manner;

in this embodiment, the preset preference index evaluation rule is obtained by performing big data statistics on a plurality of behavior indexes of the account information of the user over the entire network. When calculating the related scoring rules, each user behavior index has no correlation, so that a plurality of behavior indexes of the user account information of the whole network are separately stored and can be realized by the card storage technology, and different user behavior indexes are stored in the card message queue.

In this embodiment, a message topic is established for each user behavior index, and each message topic is divided into two message queues, for example:

time of opening an account topic: storing account opening time data;

topic of transaction record: storing transaction record data;

browsing recorded topics: storing browsing record data;

capital retention record topic: storing fund retention record data;

wind control level topic: storing wind control grade data;

whether the evaluation topic is completed or not: storing whether the evaluation judgment data is finished or not;

whether an account topic was logged off: storing whether the account has been revoked determination data;

whether a biological password topic is set: storing whether to set the biometric password determination data;

whether topic in the blacklist: whether the data is determined in the blacklist is stored.

203. Every other preset period, acquiring a corresponding user behavior index from the message queue, and counting the distribution of the user behavior index in the whole network;

in this embodiment, every preset period, the preset preference index evaluation rule needs to be updated, the same type of user behavior index data is obtained from the message queue according to multiple threads, and the distribution condition of the same type of user behavior index is calculated. The method comprises the steps of dividing various types of user behavior indexes into two types, wherein one type is that the number of users corresponding to the user behavior indexes of the same type but different numerical value ranges is counted according to preset statistical granularity to serve as a distribution current situation table, the distribution current situation of the user behavior indexes is quantified by calculating a corresponding unary regression equation, for example, the fund retention amount in a fund model factor is calculated, the user distribution situation of the amount of money 0-a, 0-2a and 0-3a is calculated by taking the amount of money a as the statistical granularity, and the unary regression equation is fitted according to the distribution situation; one is to directly count the number of user behavior indicators with different values, such as for the credit model factor, counting the number of user accounts in the blacklist and the number of user accounts not in the blacklist respectively.

204. Calculating the distribution interval of the scores of the user behavior indexes based on the distribution of the user behavior indexes in the whole network to obtain a calculation result;

in this embodiment, the calculation result is a distribution interval of the user behavior index relative to the score. The user behavior index representation mode can be divided into two modes, one mode is represented by a unitary regression equation, and the other mode is represented by user behavior indexes with different values. The method comprises the steps that the account number corresponding to unit scoring can be determined by taking the total score of distribution evaluation as a base number and dividing the total user number by the base number, the account number is substituted into a unitary regression equation to obtain a corresponding user behavior index value, the numerical distribution interval of the corresponding user behavior index is increased progressively by determining the unit scoring, if the total unit fund model factor is 20 points and the current total user number is 2000, the account number corresponding to the unit scoring is 2000/20-100, and then 100, 100x1,100x2.. 2000 is substituted into the unitary regression equation to obtain the distribution intervals of 1-2 points and 2-3 points.19-20 points of fund reserve amount; the latter directly uses the corresponding numerical value as the corresponding user behavior index interval.

205. Determining preference index evaluation rules of the user account for various products based on the calculation result, wherein the preference index evaluation rules are applied to the preference index evaluation model for counting the scores of all user behavior indexes;

in this embodiment, after determining the user behavior index corresponding to the incremental increase of the unit score, the user behavior index value of the user account is respectively input into the corresponding distribution intervals through the distribution of the user behavior index values of each fraction interval, so that the score of the user behavior index can be obtained, for example, for a fund model factor, the fund remaining amount in 15 to 16 minutes is 9.8a to 11.5a, and the fund remaining amount of a user account is 10a, so that the score of the user behavior index is 15 minutes.

206. Inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

207. screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

208. and packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

In the embodiment of the invention, the generation process of the preference index evaluation rule is introduced in detail, and the relation between the user behavior index and whether the user purchases the product or not is analyzed from different dimensions aiming at the distribution of the user behavior indexes of the whole network, wherein the analysis of the user behavior indexes is more comprehensive, the preference index evaluation rule for generating the product recommendation scheme recommended to the user is more accordant with the purchase psychology of the user from the consumption capacity to the personal quality of the user and then to the real requirement of the user, and the recommended product recommendation based on big data is naturally more accurate.

Referring to fig. 3, a third embodiment of the big data based product recommendation method according to the embodiment of the present invention includes:

301. reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

302. inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various insurance products;

303. acquiring a product transaction record, wherein the transaction record comprises transaction product information and transaction account information;

in this embodiment, product records purchased by different types of transaction accounts are acquired, and product purchase preferences of the different types of transaction accounts within an effective time are obtained through big data analysis and are used as a reference for screening a product recommendation scheme. Each user account purchases products, and a transaction account is generated by taking the purchase time as a time node, namely if the same user account purchases m products at n time nodes, n transaction accounts are generated.

304. Inputting the preference index evaluation model according to the transaction account information, and counting the scores of the user behavior indexes according to the preference index evaluation model to obtain a corresponding second preference index;

in this embodiment, the transaction account information includes the account opening time of the user account and corresponding transaction records, browsing records, fund retention records, a wind control level, whether the user has ever cancelled the account, whether a biological password is set, whether evaluation is completed, whether the user behavior indexes are in a blacklist, and the like at different time nodes of product purchase, and according to a preset preference index evaluation rule, the score of each user behavior index is calculated and accumulated to obtain a corresponding second preference index.

305. Calculating the difference value between the first preference index and the second preference index, and screening a preset number of corresponding transaction accounts in a descending order according to the difference value;

in this embodiment, the similarity between the user account and the transaction account for purchasing the product in the transaction record is quantified by calculating a difference between the first preference index and the second preference index, and the smaller the difference is, the higher the similarity between the two is, wherein the difference is not positive or negative, and the larger the absolute value is, the larger the difference is. And when the difference value of the two transaction accounts exceeds the preset difference value of the developer, discarding the transaction account which is larger than the preset difference value, and taking the rest transaction accounts as the reference transaction accounts selected by the product recommendation scheme.

In this embodiment, if the first preference index of the user account 1 is 80, the second preference index of the transaction account 1 is 76, and the second preference index of the transaction account 2 is 88, the difference between the first preference index of the user account 1 and the second preference index of the transaction account is 80-76-4, the difference between the first preference index of the user account 1 and the second preference index of the transaction account 2 is 80-88-8, and then the difference is normalized to be 8.

306. Screening the transaction product information corresponding to the transaction account to obtain a corresponding product recommendation scheme;

in this embodiment, the products purchased by each transaction account can be obtained according to the historical transaction records, and then the products of the user account preference similar to the transaction accounts are analyzed, so as to determine the products corresponding to the target user account.

In this embodiment, for the screened transaction accounts, the purchase quantity of each transaction product information is counted, and the purchase quantities are sorted by at least one, and transaction product information of a preset quantity is screened as and written in an initial product recommendation scheme list as an initial product recommendation scheme, wherein if the transaction product information quantity is less than the preset quantity, the transaction product information of the transaction product information quantity is taken.

307. And packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

In the embodiment of the invention, the second preference index of each user is calculated through the past product purchase records, then the users who purchase products with higher similarity with the current user are selected according to the first preference indexes of the users, the products purchased by the past users are selected and recommended to the current user, and the generated product recommendation scheme is more suitable for the current user according to the history.

Referring to fig. 4, a fourth embodiment of the big data based product recommendation method according to the embodiment of the present invention includes:

401. reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

402. inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

403. screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

404. packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account;

405. monitoring whether a client corresponding to the user account sends feedback information corresponding to the product recommendation scheme;

in this embodiment, after the user obtains the product recommendation scheme, the user may select to purchase a product in the scheme, or evaluate the product recommendation accuracy based on the big data in the scheme, so that the server automatically adjusts the product recommendation ranking or addition and deletion based on the big data in the recommendation scheme. Wherein, the server determines whether the front end sends the feedback information by monitoring the command.

406. If yes, obtaining the feedback information, and extracting characteristic information in the feedback information;

in this embodiment, the feedback information carries opinion feedback of the user on the product recommendation scheme, and the feedback is performed according to characteristic information in the feedback information. The characteristic information includes: the user purchases the corresponding product, indicated by the character "1"; (ii) a The feedback opinions of each product recommendation scheme are represented in the form of product + feedback opinions, if no feedback opinions exist, the feedback opinions are represented by characters of '0', if recommendation is wrong, the feedback opinions are represented by characters of '1', and if recommendation is correct, the feedback opinions are represented by characters of '2'.

407. Determining a correction value corresponding to the product recommendation scheme based on the characteristic information;

in this embodiment, each type of feature information has a correction value corresponding to modification of a product recommendation scheme, and the correction value is used for adjusting the ordering order of products in product recommendation schemes corresponding to the user account and other user accounts similar to the user account. If the user purchases the corresponding product, the position of the product in the corresponding recommended scheme is moved forward by a preset amount in the sequence; if the user feeds back a recommendation error to a certain product in the product recommendation scheme, removing the product in the product recommendation scheme corresponding to the user account, and moving the product in the product recommendation scheme corresponding to the similar user account by a preset number of positions after sequencing; and if the user feeds back and recommends a product in the product recommendation scheme correctly, the product in the product recommendation scheme corresponding to the user account is set at the top, and the product is moved forward by a plurality of positions in the product recommendation scheme corresponding to the similar user account.

408. And adjusting the product recommendation scheme corresponding to the user account based on the correction value.

In this embodiment, the correction values determine specific contents of adjustment of the product recommendation scheme, traverse all the correction values, and adjust corresponding products in the corresponding product recommendation scheme according to the sequence.

In the embodiment of the invention, whether the recommended products conform to the purchase psychology of the user can be evaluated through the feedback information of the user on the product recommendation scheme, so that the product recommendation scheme of the users of the same type can be adjusted, and the recommendation accuracy is optimized.

In the above description of the product recommendation method based on big data in the embodiment of the present invention, referring to fig. 5, a product recommendation device based on big data in the embodiment of the present invention is described below, where an embodiment of the product recommendation device based on big data in the embodiment of the present invention includes:

the reading module 501 is configured to read user account information of a product to be recommended, where the user account information includes a plurality of user behavior indexes;

the index evaluation module 502 is used for inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

a scheme generating module 503, configured to filter products corresponding to the user account based on the first preference index, to obtain a corresponding product recommendation scheme;

and the scheme pushing module 504 is configured to encapsulate the product recommendation scheme and push the product recommendation scheme to a corresponding user account.

In the embodiment of the invention, a plurality of behavior indexes of a user account are used as evaluation parameters of a product recommendation scheme, the parameters of the behavior indexes are counted to obtain the product purchase preference of the account, a preset number of matched products are screened according to the quantified account purchase preference to combine the product recommendation scheme of the account, and then the product recommendation scheme is packaged for displaying the product recommendation scheme when the account logs in.

Referring to fig. 6, another embodiment of the big data based product recommendation apparatus according to the embodiment of the present invention includes:

the reading module 601 is configured to read user account information of a product to be recommended, where the user account information includes a plurality of user behavior indexes;

the index evaluation module 602 is configured to input the user account information into a preset preference index evaluation model, and count scores of the multiple user behavior indexes through the preference index evaluation model to obtain a first preference index of the corresponding user account for satisfaction of various products;

a scheme generating module 603, configured to filter, based on the first preference index, a product corresponding to the user account to obtain a corresponding product recommendation scheme;

and the scheme pushing module 604 is configured to encapsulate the product recommendation scheme and push the product recommendation scheme to a corresponding user account.

Specifically, the big data based product recommendation device further includes:

an obtaining module 605, configured to obtain account information of a user in a whole network and store the account information in a message queue in a distributed manner;

a counting module 606, configured to obtain, every preset period, a corresponding user behavior index from the message queue, and count distribution of the user behavior index in the whole network;

a calculating module 607, configured to calculate a distribution interval of scores of the user behavior indicators based on distribution of the user behavior indicators in the whole network, so as to obtain a calculation result;

an evaluation rule generating module 608, configured to determine, based on the calculation result, a preference index evaluation rule of the user account for each product type, where the preference index evaluation rule is applied to the preference index evaluation model for counting scores of each user behavior index.

Specifically, the preference index evaluation model comprises a plurality of evaluation indexes, wherein the evaluation indexes comprise an account opening model factor, a transaction model factor, a browsing model factor, a fund model factor, an evaluation model factor, a logout model factor, an encryption model factor, a credit model factor and a wind control model factor;

wherein the preference index evaluation rule comprises:

determining the score of the account opening model factor according to the account opening time, wherein the longer the account opening time is, the lower the score of the account opening model factor is; or

Determining the grade of the transaction model factor according to the transaction time and the transaction amount of the account, wherein the closer the transaction time is, the larger the transaction amount is, and the higher the grade of the transaction model factor is; or

Determining the score of the browsing model factor according to the time length and the number of the products browsed by the account, wherein the score of the browsing model factor is higher when the time length of browsing the products is larger and the number of the products is larger; or

Determining the score of the fund model factor according to the account fund remaining quantity and the remaining time, wherein the larger the remaining quantity is, the longer the remaining time is, the higher the score of the fund model factor is; or

If the user account is evaluated, the evaluation model factor is fully scored, otherwise, the score is not added; or

If the account has no log-off record, logging off the model factor to obtain full score, otherwise, not adding score; or

If the user account is encrypted by biological identification, the encryption model factor is fully scored, otherwise, the score is not added;

if the account is not on the blacklist, the credit model factor is fully scored, otherwise, the score is not added; or

And determining the score of the wind control model factor according to the wind control level of the account, wherein the less the wind control level is, the higher the score of the wind control model factor is.

It is noted that the preference index evaluation rule includes one or more of the above rules.

Specifically, the scheme generating module 603 further includes:

an obtaining unit 6031, configured to obtain a product transaction record, where the transaction record includes transaction product information and transaction account information;

a statistic unit 6032, configured to input the preference index evaluation model according to the transaction account information, and count scores of the multiple user behavior indexes according to the preference index evaluation model to obtain corresponding second preference indexes;

a calculating unit 6033, configured to calculate a difference between the first preference index and the second preference index, and filter a preset number of corresponding transaction accounts in a descending order according to the difference;

and the generating unit 6034 is configured to filter the transaction product information corresponding to the transaction account to obtain a corresponding product recommendation scheme.

Specifically, the big data based product recommendation device further includes:

the monitoring module 609 is configured to monitor whether the client corresponding to the user account uploads feedback information corresponding to the product recommendation scheme;

the extracting module 610 is configured to obtain feedback information corresponding to the product recommendation scheme if the feedback information is uploaded by the client corresponding to the user account, and extract feature information in the feedback information;

the correction module 611 is configured to determine, based on the feature information, a correction value corresponding to the product recommendation scheme; and adjusting the product recommendation scheme corresponding to the user account based on the correction value.

Specifically, the product recommendation device further includes an association module 612, where the association module 612 is specifically configured to:

determining a family member structure corresponding to the user account according to the user account information;

judging whether a family associated account corresponding to the family member structure exists or not;

if so, acquiring a product recommendation scheme corresponding to the family associated account;

and screening a preset number of products according to the product recommendation sequence of the product recommendation scheme corresponding to the family associated account, and randomly adding the products into the product recommendation scheme corresponding to the user account.

Specifically, the product recommendation device further includes an identification module 613, where the identification module 613 is specifically configured to:

counting the grade of each evaluation index in the user account, and selecting the evaluation index with the highest grade as the advantage evaluation index of the user account;

screening a preset number of products as dominant recommended products of the user account according to the product recommendation sequence of the product recommendation scheme;

and determining a corresponding product label according to the advantage evaluation index and the advantage recommended product so as to identify the corresponding advantage recommended product when the product recommendation scheme is pushed.

In the embodiment of the invention, a generation process of a preference index evaluation rule is introduced in detail by taking a plurality of behavior indexes of a user account as evaluation parameters of a product recommendation scheme, and the relationship between the user behavior indexes and whether a user purchases a product is analyzed from different dimensions aiming at the distribution of the user behavior indexes of the whole network, wherein the analysis of the user behavior indexes is more comprehensive, and the preference index evaluation rule for recommending the product recommendation scheme to the user is generated to be more in line with the purchase psychology of the user from the consumption capacity to the personal quality of the user and then to the real requirement of the user; calculating a second preference index of each user according to the previous product purchase record, selecting users who purchase products with higher similarity with the current user according to the first preference index of the users, and selecting the products purchased by the past users and recommending the products to the current user according to the history; and finally, whether the recommended products accord with the purchase psychology of the users can be evaluated through the feedback information of the users to the product recommendation scheme, so that the product recommendation scheme of the users of the same type is adjusted, and the recommendation accuracy is optimized.

Fig. 5 and 6 describe the big data based product recommendation device in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the big data based product recommendation device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

Fig. 7 is a schematic structural diagram of a big data based product recommendation device 700 according to an embodiment of the present invention, where the big data based product recommendation device 700 may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 710 (e.g., one or more processors) and a memory 720, one or more storage media 730 (e.g., one or more mass storage devices) storing an application 733 or data 732. Memory 720 and storage medium 730 may be, among other things, transient storage or persistent storage. The program stored in the storage medium 730 may include one or more modules (not shown), each of which may include a series of instruction operations for the big data based product recommendation device 700. Still further, the processor 710 may be configured to communicate with the storage medium 730 to execute a series of instruction operations in the storage medium 730 on the big data based product recommendation device 700.

The big data based product recommendation device 700 may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input-output interfaces 760, and/or one or more operating systems 731, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the configuration of the big data based product recommendation device illustrated in FIG. 7 does not constitute a limitation of the big data based product recommendation device, and may include more or less components than those illustrated, or some components in combination, or a different arrangement of components.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and may also be a volatile computer-readable storage medium, having stored therein instructions, which, when executed on a computer, cause the computer to perform the steps of the big data-based product recommendation method.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses, and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A big data based product recommendation method is characterized by comprising the following steps:

reading user account information of a product to be recommended, wherein the user account information comprises a plurality of user behavior indexes;

inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

and packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

2. The big data-based product recommendation method according to claim 1, wherein before the inputting the user account information into a preset preference index evaluation model and counting the scores of the plurality of user behavior indexes through the preference index evaluation model to obtain a first preference index corresponding to the satisfaction degree of the user account with each product, the method further comprises:

acquiring account information of a whole network user and storing the account information in a message queue in a distributed manner;

every other preset period, acquiring a corresponding user behavior index from the message queue, and counting the distribution of the user behavior index in the whole network;

calculating the distribution interval of the scores of the user behavior indexes based on the distribution of the user behavior indexes in the whole network to obtain a calculation result;

and determining a preference index evaluation rule of the user account for various products based on the calculation result, wherein the preference index evaluation rule is applied to the preference index evaluation model for counting the scores of the user behavior indexes.

3. The big-data based product recommendation method according to claim 2, wherein the preference index evaluation model comprises a plurality of evaluation indexes, the evaluation indexes comprising an account opening model factor, a transaction model factor, a browsing model factor, a fund model factor, an evaluation model factor, a logout model factor, an encryption model factor, a credit model factor, a wind control model factor;

wherein the preference index evaluation rule comprises:

determining the score of the account opening model factor according to the account opening time, wherein the longer the account opening time is, the lower the score of the account opening model factor is; or

Determining the grade of the transaction model factor according to the transaction time and the transaction amount of the account, wherein the closer the transaction time is, the larger the transaction amount is, and the higher the grade of the transaction model factor is; or

Determining the score of the browsing model factor according to the time length and the number of the products browsed by the account, wherein the score of the browsing model factor is higher when the time length of browsing the products is larger and the number of the products is larger; or

Determining the score of the fund model factor according to the account fund remaining quantity and the remaining time, wherein the larger the remaining quantity is, the longer the remaining time is, the higher the score of the fund model factor is; or

If the user account is evaluated, the evaluation model factor is fully scored, otherwise, the score is not added; or

If the account has no log-off record, logging off the model factor to obtain full score, otherwise, not adding score; or

If the user account is encrypted by biological identification, the encryption model factor is fully scored, otherwise, the score is not added;

if the account is not on the blacklist, the credit model factor is fully scored, otherwise, the score is not added; or

And determining the score of the wind control model factor according to the wind control level of the account, wherein the less the wind control level is, the higher the score of the wind control model factor is.

4. The big-data-based product recommendation method according to claim 2, wherein the screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme comprises:

acquiring a product transaction record, wherein the transaction record comprises transaction product information and transaction account information;

inputting the preference index evaluation model according to the transaction account information, and counting the scores of the user behavior indexes according to the preference index evaluation model to obtain a corresponding second preference index;

calculating the difference value between the first preference index and the second preference index, and screening a preset number of corresponding transaction accounts in a descending order according to the difference value;

and screening the transaction product information corresponding to the transaction account to obtain a corresponding product recommendation scheme.

5. The big-data-based product recommendation method according to claim 4, further comprising, after packaging the product recommendation scheme and pushing it to a corresponding user account:

monitoring whether a client corresponding to the user account uploads feedback information corresponding to the product recommendation scheme;

if yes, obtaining the feedback information, and extracting characteristic information in the feedback information;

determining a correction value corresponding to the product recommendation scheme based on the characteristic information;

and adjusting the product recommendation scheme corresponding to the user account based on the correction value.

6. The big data based product recommendation method according to any one of claims 1-5, further comprising:

determining a family member structure corresponding to the user account according to the user account information;

judging whether a family associated account corresponding to the family member structure exists or not;

if so, acquiring a product recommendation scheme corresponding to the family associated account;

and screening a preset number of products according to the product recommendation sequence of the product recommendation scheme corresponding to the family associated account, and randomly adding the products into the product recommendation scheme corresponding to the user account.

7. The big-data-based product recommendation method according to claim 6, wherein after the screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme, the method further comprises:

counting the scores of all evaluation indexes in the user account, and selecting the evaluation index with the highest score as the advantage evaluation index of the user account;

screening a preset number of products as dominant recommended products of the user account according to the product recommendation sequence of the product recommendation scheme;

and determining a corresponding product label according to the advantage evaluation index and the advantage recommended product so as to identify the corresponding advantage recommended product when the product recommendation scheme is pushed.

8. A big-data based product recommendation device, comprising:

the system comprises a reading module, a recommendation module and a recommendation module, wherein the reading module is used for reading user account information of a product to be recommended, and the user account information comprises a plurality of user behavior indexes;

the index evaluation module is used for inputting the user account information into a preset preference index evaluation model, and counting the scores of the user behavior indexes through the preference index evaluation model to obtain a first preference index of the satisfaction degree of the corresponding user account to various products;

the scheme generation module is used for screening products corresponding to the user account based on the first preference index to obtain a corresponding product recommendation scheme;

and the scheme pushing module is used for packaging the product recommendation scheme and pushing the product recommendation scheme to a corresponding user account.

9. A big-data-based product recommendation device, comprising: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;

the at least one processor invokes the instructions in the memory to cause the big-data based product recommendation device to perform the big-data based product recommendation method of any of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the big-data based product recommendation method according to any one of claims 1-7.

Images