CN114782062A - Commodity recall optimization method and device, equipment, medium and product thereof - Google Patents
Commodity recall optimization method and device, equipment, medium and product thereof Download PDFInfo
- Publication number
- CN114782062A CN114782062A CN202210555827.3A CN202210555827A CN114782062A CN 114782062 A CN114782062 A CN 114782062A CN 202210555827 A CN202210555827 A CN 202210555827A CN 114782062 A CN114782062 A CN 114782062A
- Authority
- CN
- China
- Prior art keywords
- recall
- commodity
- commodities
- source
- score
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/01—Customer relationship services
- G06Q30/014—Providing recall services for goods or products
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a commodity recall optimization method, a device, equipment, a medium and a product thereof, wherein the method comprises the following steps: obtaining a plurality of preset commodity data subsets recalled by a plurality of recall sources correspondingly, wherein each subset comprises recalled commodities and a matching degree representing the recalled commodities; determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset; respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source; and merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity. According to the method and the device, the information contribution degree of the commodities obtained by different recall sources can be quantized in fine granularity, so that a more accurate final recommendation result can be determined according to the actual scores of the commodities, and the method and the device are suitable for being used by independent sites.
Description
Technical Field
The present application relates to the field of e-commerce information technologies, and in particular, to a method for optimizing a commodity recall, and a corresponding apparatus, computer device, computer-readable storage medium, and computer program product.
Background
The method for recommending the commodities in the e-commerce platform is a high-frequency application, is widely used for application scenes such as commodity searching, commodity advertisement putting, commodity screening and the like, can improve commodity matching efficiency, enables the commodities of merchant users to be sold more easily, and enables the requirements of consumer users to be met more easily.
When the commodity recommendation algorithm is applied, one key link is a commodity recalling link. In the commodity recalling link, corresponding commodities are obtained by different strategies or through different data source channels by starting a plurality of preset recalling sources. The recalled commodities can be used as results according to actual conditions, and can also be further used in a sequencing mode.
The quantity of commodities obtained from each recall source is large, and the information contribution value corresponding to each commodity is not unified and quantized, so that commodities obtained from different recall sources are integrated into a final result to be used, and the final obtained result commodity is difficult to achieve expectation.
For such problems, the conventional processing method adjusts the importance of commodities in different recall sources by simply adjusting the weight of each recall source or matching a neural network model based on deep learning, but since each recall source has multiple commodities, the number of commodities expected to be obtained finally is often small, so that the method has very limited practical functions due to coarse information processing granularity, and when the neural network model is used for implementation, the training cost is high, and the method is difficult to achieve effects for on-line stores deployed as independent sites.
In view of this, the applicant develops a new way to explore a new idea corresponding to the realization of commodity recall optimization, and then draws the present application.
Disclosure of Invention
It is an object of the present application to solve the above-mentioned problems and provide a method for product recall optimization and corresponding apparatus, computer device, computer readable storage medium, computer program product,
The technical scheme is adopted to adapt to various purposes of the application as follows:
in one aspect, a method for product recall optimization is provided, adapted for one of the purposes of the present application, comprising the steps of:
obtaining a plurality of preset commodity data subsets recalled correspondingly by a plurality of recall sources, wherein each subset comprises recalled commodities and a matching degree representing the recalled commodities;
determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset;
respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source;
and merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity.
Optionally, the determining, according to the user behavior data corresponding to the commodities in each commodity data subset, a plurality of evaluation indexes of each recall source includes the following steps:
acquiring user behavior data corresponding to commodities in the commodity data subset corresponding to each recall source;
determining parameter values corresponding to all evaluation indexes of each recall source according to the user behavior data;
and calculating the evaluation indexes corresponding to the recall sources by applying the preset algorithm corresponding to each evaluation index and the corresponding parameter values thereof.
Optionally, the step of respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source includes the following steps:
sorting the recall sources according to each evaluation index to obtain a recall source sorting list corresponding to each evaluation index;
according to the numerical value of the evaluation index, a uniform ranking score sequence is applied, and corresponding ranking scores from high to low are set for the recall sources in each recall source ranking list;
and matching the preset weights with the sorting scores of the recall source sorting list corresponding to each evaluation index for summation to obtain the index score corresponding to each recall source.
Optionally, before the step of obtaining a plurality of product data subsets recalled correspondingly by a plurality of preset recall sources, the method includes the following steps:
and responding to the commodity matching instruction, calling a plurality of recall sources, and matching out a corresponding commodity data subset for the target commodity specified by the commodity matching instruction based on each recall source, wherein the commodities in the commodity data subset are similar to the target commodity in composition characteristics.
Optionally, after the step of merging the commodity data subsets into a commodity recommendation set, the method includes the following steps:
performing reverse ordering on the commodity recommendation set according to the actual scores of the commodities in the commodity recommendation set;
obtaining a preset plurality of commodities which are ranked at the front from the commodity recommendation set to form a commodity recommendation list;
and pushing the commodity recommendation list to target terminal equipment.
Optionally, the evaluation index includes any one or more of the following items: the system comprises a recall rate, an accuracy rate, a reconciliation score, an area under a curve and an area under a user curve, wherein the reconciliation score is determined according to the recall rate and the accuracy rate.
In another aspect, an object of the present application is to provide a product recall optimization apparatus, which includes a data obtaining module, an index determining module, an index summarizing module, and a data merging module, wherein: the data acquisition module is used for acquiring a plurality of preset commodity data subsets recalled by a plurality of recall sources correspondingly, and each subset comprises recalled commodities and a matching degree representing the recalled commodities; the index determining module is used for determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset; the index summarizing module is used for summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source; and the data merging module is used for merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity.
Optionally, the index determining module includes: the behavior acquisition unit is used for acquiring user behavior data corresponding to the commodities in the commodity data subset corresponding to each recall source; the parameter determining unit is used for determining parameter values corresponding to the evaluation indexes of each recall source according to the user behavior data; and the index calculation unit is used for applying the preset algorithm corresponding to each evaluation index and the corresponding parameter value thereof to calculate the evaluation index corresponding to each recall source.
Optionally, the index summarizing module includes: the sequencing processing unit is used for sequencing all the recall sources according to each evaluation index to obtain a recall source sequencing list corresponding to each evaluation index; the sequencing assignment unit is used for applying a uniform sequencing score sequence according to the numerical value of the evaluation index and setting a corresponding sequencing score from high to low for the recall sources in each recall source sequencing list; and the summing and summarizing unit is used for matching preset weights with the sorting scores of the recall source sorting list corresponding to each evaluation index for summation to obtain the index score corresponding to each recall source.
Optionally, prior to the data obtaining module, the data obtaining module includes: and the recall execution module is used for responding to the commodity matching instruction, calling a plurality of recall sources, and matching out a corresponding commodity data subset for the target commodity specified by the commodity matching instruction based on each recall source, wherein the commodities in the commodity data subset have similar composition characteristics with the target commodity.
Optionally, the data merging module further includes: the merging and sequencing module is used for performing reverse sequencing on the commodity recommendation set according to the actual scores of the commodities in the commodity recommendation set; the list generation module is used for acquiring a preset plurality of commodities which are ranked at the front from the commodity recommendation set to form a commodity recommendation list; and the list pushing module is used for pushing the commodity recommendation list to target terminal equipment.
Optionally, the evaluation index includes any one or more of the following items: the system comprises a recall rate, an accuracy rate, a reconciliation score, an area under a curve and an area under a user curve, wherein the reconciliation score is determined according to the recall rate and the accuracy rate.
In yet another aspect, a computer apparatus adapted to one of the purposes of the present application is provided, and comprises a central processing unit and a memory, wherein the central processing unit is used for calling and running a computer program stored in the memory to execute the steps of the product recall optimization method described in the present application.
In another aspect, a computer-readable storage medium is provided, which stores a computer program implemented according to the product recall optimization method in the form of computer-readable instructions, and when the computer program is called by a computer, executes the steps included in the method.
In yet another aspect, a computer program product is provided that is adapted to carry out another object of the present application and includes computer program/instructions which, when executed by a processor, implement the steps of the method for merchandise recall optimization described in any of the embodiments of the present application.
Compared with the prior art, the application has various advantages, at least comprising: according to the method and the device, a plurality of evaluation indexes are respectively determined through commodity data subsets obtained for a plurality of recall sources, index scores corresponding to each recall source are determined according to the evaluation indexes of each recall source, actual scores of commodities are jointly determined according to the index scores and the matching degree of the commodities, and the fine granularity quantization of the information contribution degree of the commodities obtained by different recall sources is realized, so that a more accurate final recommendation result can be determined according to the actual scores of the commodities, a complex mathematical model is not needed in the implementation process of the method and the device, the method and the device can be realized with lower system overhead, the deployment cost is lower, and the method and the device are particularly suitable for independent sites in an e-commerce platform.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a schematic flow chart of an exemplary embodiment of a product recall optimization method according to the present application.
Fig. 2 is a flowchart illustrating a process of determining each evaluation index of each recall source in an embodiment of the present application.
FIG. 3 is a flowchart illustrating a process of determining an indicator score for each recall source in an embodiment of the present application.
Fig. 4 is a flowchart illustrating a process of obtaining a product recommendation list according to a product recommendation set in an embodiment of the present application.
FIG. 5 is a functional block diagram of a merchandise recall optimization apparatus according to the present application;
fig. 6 is a schematic structural diagram of a computer device used in the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary only for the purpose of explaining the present application and are not to be construed as limiting the present application.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.
It will be understood by those within the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
As used herein, "client," "terminal," and "terminal device" include both wireless signal receiver devices, which are only capable of wireless signal receiver devices without transmit capability, and receiving and transmitting hardware devices, which have receiving and transmitting hardware capable of two-way communication over a two-way communication link, as will be understood by those skilled in the art. Such a device may include: cellular or other communication devices such as personal computers, tablets, etc. having a single line display or a multi-line display or cellular or other communication devices without a multi-line display; PCS (Personal Communications Service), which may combine voice, data processing, facsimile and/or data communication capabilities; a PDA (Personal Digital Assistant) that may include a radio frequency receiver, a pager, internet/intranet access, web browser, notepad, calendar, and/or GPS (Global Positioning System) receiver; a conventional laptop and/or palmtop computer or other device having and/or including a radio frequency receiver. As used herein, a "client," "terminal device" can be portable, transportable, installed in a vehicle (aeronautical, maritime, and/or land-based), or situated and/or configured to operate locally and/or in a distributed fashion at any other location(s) on earth and/or in space. The "client", "terminal Device" used herein may also be a communication terminal, a web terminal, a music/video playing terminal, such as a PDA, an MID (Mobile Internet Device) and/or a Mobile phone with music/video playing function, and may also be a smart tv, a set-top box, and the like.
The hardware referred to by the names "server", "client", "service node", etc. in the present application is essentially an electronic device with the performance of a personal computer, and is a hardware device having necessary components disclosed by the von neumann principles such as a central processing unit (including an arithmetic unit and a controller), a memory, an input device, and an output device, in which a computer program is stored in the memory, and the central processing unit loads a program stored in an external memory into the internal memory to run, executes instructions in the program, and interacts with the input and output devices, thereby accomplishing specific functions.
It should be noted that the concept of "server" in the present application can be extended to the case of server cluster. According to the network deployment principle understood by those skilled in the art, the servers should be logically divided, and in physical space, the servers may be independent from each other but can be called through an interface, or may be integrated into one physical computer or a set of computer clusters. Those skilled in the art will appreciate this variation and should not be so limited as to restrict the implementation of the network deployment of the present application.
One or more technical features of the present application, unless expressly specified otherwise, may be deployed to a server for implementation by a client remotely invoking an online service interface provided by a capture server for access, or may be deployed directly and run on the client for access.
Unless specified in clear text, the neural network model referred to or possibly referred to in the application can be deployed in a remote server and performs remote invocation at a client, and can also be deployed in a client with sufficient equipment capability to perform direct invocation.
Various data referred to in the present application may be stored in a server remotely or in a local terminal device unless specified in the clear text, as long as the data is suitable for being called by the technical solution of the present application.
The person skilled in the art will know this: although the various methods of the present application are described based on the same concept so as to be common to each other, they may be independently performed unless otherwise specified. In the same way, for each embodiment disclosed in the present application, the same inventive concept is proposed, and therefore, concepts expressed in the same manner and concepts expressed in terms of the same are equally understood, and even though the concepts are expressed differently, they are merely convenient and appropriately changed.
The embodiments to be disclosed herein can be flexibly constructed by cross-linking related technical features of the embodiments unless the mutual exclusion relationship between the related technical features is stated in the clear text, as long as the combination does not depart from the inventive spirit of the present application and can meet the needs of the prior art or solve the deficiencies of the prior art. Those skilled in the art will appreciate variations therefrom.
The commodity recall optimization method can be programmed into a computer program product and deployed in a client or a server to be operated, for example, in an exemplary application scenario of the application, the commodity recall optimization method can be deployed and implemented in the server of an e-commerce platform, so that the method can be executed by accessing an interface opened after the computer program product is operated and performing human-computer interaction with a process of the computer program product through a graphical user interface.
Referring to fig. 1, the merchandise recall optimization method of the present application, in an exemplary embodiment thereof, includes the following steps:
step S1100, obtaining a plurality of preset commodity data subsets recalled by a plurality of recall sources correspondingly, wherein each subset comprises recalled commodities and a matching degree representing the recalled commodities;
and a plurality of recall sources are configured in the e-commerce platform, each recall source provides a commodity data recall service corresponding to a recall strategy, and the commodity data is matched from a corresponding recall channel to obtain a corresponding commodity data subset. The merchandise data subset typically includes a plurality of merchandise and a degree of match obtained by each merchandise under a corresponding recall source to match the recall source.
In an application scenario of multi-path recall based on target commodities, a target commodity is given in advance, a plurality of recall sources for realizing similar matching are called, wherein a first recall source performs similar matching with commodity pictures and/or deep semantic feature information of commodity description information of each commodity in a first commodity data pool according to deep semantic feature information of the commodity pictures and/or the commodity description information of the target commodity, the matching degree of the commodities which form matching with the target commodity is correspondingly calculated, and then commodities with the matching degree higher than a preset threshold value are extracted from the first commodity data pool to form a first commodity data subset corresponding to the first recall source. The second recall source is the same as the first recall source, similar matching is carried out on the deep semantic feature information of the commodity picture and/or the commodity description information of each commodity in the second commodity data pool according to the deep semantic feature information of the commodity picture and/or the commodity description information of the target commodity, the matching degree of the commodities which form matching with the target commodity is correspondingly calculated, and then the commodities of which the matching degree is higher than a preset threshold value are extracted from the second commodity data pool to form a second commodity data subset corresponding to the first recall source. The first commodity data pool may store high profit commodities, and the second commodity data pool may store hot sales commodities.
It is understood from the above examples that those skilled in the art may preset the recall sources, preset corresponding recall policies and/or recall channels for each recall source, implement corresponding recall services, provide corresponding recall interfaces, call corresponding recall sources when commodity data recall is required, so as to obtain corresponding commodity data subsets thereof, and call multiple recall sources concurrently, so as to obtain multiple commodity data subsets accordingly. Each commodity data subset is acquired based on a different recall strategy and/or recall channel, so that the commodities between the commodity data subsets may be individually identical, and for the same commodities, preference may be subsequently made at the stage of merging.
Step S1200, determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset;
the practical effectiveness of each recall source may be analyzed by the user behavior data generated by the access of the respective commodity by the user for the respective subset of commodity data correspondingly obtained by the respective recall source.
The user behavior data can be obtained by embedding points in a commodity display page of a certain commodity, when any user accesses one commodity display page, the corresponding user behavior data is submitted to the server through the corresponding embedded point code of the page, the user is represented to execute corresponding operation on the corresponding commodity aiming at the commodity display page, and a corresponding operation event is generated. The user behavior data is stored in a log database in the form of logs, and the user behavior data of the commodities within a preset certain time range can be directly called from the log database according to the commodities recalled from each recall source, so that the evaluation indexes of each recall source can be determined.
By counting the user behavior data of the commodities in the commodity data subset obtained by each recall source, each evaluation index corresponding to each recall source can be obtained. The evaluation index comprises any one or more of the following items: recall, accuracy, harmonic score, area under the curve, area under the user curve, may be selected by those skilled in the art as desired. As an example, the present embodiment may employ the above respective evaluation indexes at the same time for realizing the evaluation of the recall performance of the respective recall sources.
The accuracy PrecisionSingleIn the present application, the number of commodities Click visited by the user in each corresponding commodity data subset of the recall source is definedSingleAnd the total Amount of commodities Amount in the commodity data subsetSingleThe ratio of (a) to (b). The formula is as follows:
the Recall rate RecallSingleIn this application, it is defined as corresponding to each recall sourceNumber of commodities Click visited by user in commodity data subsetSingleTotal Amount of commodities Amount occupying each commodity data subset corresponding to all recalling sourcesSingleThe ratio of the sum of (a) and (b). The formula is expressed as:
the harmony score FSingleThe method is a harmonic average of the accuracy rate and the recall rate and is used for balancing the accuracy rate and the recall rate, the accuracy rate and the recall rate are generally in negative correlation, and the accuracy rate and the recall rate are generally contradictory, so that the influence of the accuracy rate and the recall rate can be balanced by introducing a harmonic score, and the higher the harmonic score is, the higher the model quality is. The harmonic score may be calculated according to the following formula:
wherein, beta can be flexibly adjusted according to the requirement on the importance of the recall rate, and when the beta is equal to 1, the accuracy rate and the recall rate are equally important; when the accuracy rate is less than 1, the accuracy rate is heavier than the recall rate; when it is greater than 1, it means that the recall rate is heavier than the accuracy rate. Since the present application focuses on recalling optimizations, the exemplary setting β ═ 1.
The Area under the curve (AUC) is one of the indexes for evaluating the quality of a recall source, and is defined by the following formula in the application:
wherein, unClickSingleThe number of items in the corresponding recall source that were not accessed by the user. Thus, it can be negatively charged that the area under the curve reflects one ordering capability among the entire sample of recall source recalls.
The area Gauc under the user curve is a result obtained by calculating the area under the curve based on each user. In the field of advertisement calculation, what is actually measured is the sorting capacity of different users for different advertisement commodities, so that the sorting capacity of the same user for different advertisements is more concerned actually, and the area under the curve can be calculated based on a single user. To this end, the following formula is applied:
wherein (u, p) represents user and commodity
According to the above formula: gauc (group Auc) actually calculates Auc of each user, then weights and averages, and finally obtains Gauc, so that the influence of poor comparison of ordering results among different users can be reduced. In actual processing, the weight w(u,p)It can be generally set as the number of times each user accesses.
From the above disclosure of exemplary individual evaluation index formulas, it can be understood that, according to one or more preselected evaluation index formulas, when it is necessary to evaluate each recall source based on the commodity data subset corresponding to each recall source, the corresponding individual evaluation indexes can be calculated by applying the corresponding evaluation index formulas according to the relevant parameter values determined according to the user behavior data.
Step S1300, respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source;
for each recall source, the evaluation indexes are relatively dispersed, the evaluation information of the corresponding recall source is provided from different angles, and the method is not intuitive, so that a preset summarizing algorithm can be applied to synthesize a plurality of evaluation indexes corresponding to each recall source into a corresponding single index score for comprehensively indicating the quality of each recall source. In one embodiment, the above evaluation indexes are directly added to obtain corresponding index scores. In another embodiment, the matching weights may also be summarized for each evaluation index, for example, an exemplary formula is as follows:
Score＝w1*PrecisionSingle+w2*RecallSingle+w3*FSingle+w4*AucSingle+w5*GaucSingle
wherein w is a predetermined weight, which can be predetermined by those skilled in the art.
The method has the advantages that the evaluation indexes corresponding to the recall sources are gathered to determine the scores of the indexes, so that the advantages and the disadvantages of the recall sources are comprehensively evaluated from different dimensions, the overall performance of the recall sources is unified into the same dimension, and the method is convenient for follow-up guidance to sort the full-scale commodities recalled from all the recall sources.
In addition to the above, other manners for determining the index score may be adopted, and the following embodiments of the present application will further disclose implementation of other manners, which are not shown here.
And step S1400, merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity.
In order to realize the summarization of commodities in the commodity data subsets corresponding to all recall sources, all commodity data subsets can be combined to obtain a commodity recommendation set. In the merging process, for the condition that different commodity data subsets have the same commodity, only the data records corresponding to the recall sources with higher index scores can be reserved, and duplicate removal is realized. The obtained commodity recommendation sets can be directly used as result sets through combination, and can also be pushed to a preset sequencing model or applied with a preset sequencing algorithm, and then output after sequencing, and the method can be implemented by the technical personnel in the field as required.
In order to facilitate uniform sorting of the commodities in the commodity recommendation set, the matching degree of each commodity obtained from the corresponding recall source and the index score obtained from the recall source can be multiplied to obtain a product, the product is used as an actual score corresponding to the commodity, and sorting of the commodity recommendation set can be achieved according to the actual score.
As can be readily appreciated from the above embodiments, the present application has various advantages over the prior art, including at least: according to the method and the device, a plurality of evaluation indexes are respectively determined through the commodity data subsets obtained for a plurality of recall sources, the index score corresponding to each recall source is determined according to the plurality of evaluation indexes of each recall source, the actual score of the commodity is determined jointly according to the matching degree of the index score and the commodity, and the fine granularity quantization of the information contribution degree of the commodity obtained by different recall sources is realized, so that a more accurate final recommendation result can be determined according to the actual score of the commodity, a complex mathematical model is not needed in the implementation process, the method and the device can be realized with lower system overhead, the deployment cost is lower, and the method and the device are particularly suitable for being used by independent sites in an e-commerce platform.
Based on any of the above embodiments, referring to fig. 2, in step S1200, determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to the commodities in each commodity data subset includes the following steps:
step S1210, user behavior data corresponding to the commodities in the commodity data subset corresponding to each recall source is obtained;
as previously described, for each recall source, raw user behavior data is obtained from the log database over a time range. And then, performing data cleaning on the original user behavior data, so that the user behavior data can be formatted into a unified expression, wherein the form of the unified expression is as follows:
user, current commodity, recalled commodity, degree of match, recalled source, access status
The method comprises the steps that a user indicates to execute an active user in an E-commerce platform, a current commodity indicates a target commodity used for triggering a plurality of recall sources to recall, a recall commodity indicates that the recall source recalls a corresponding commodity according to the target commodity, the recall source indicates that the recall source corresponding to the recall commodity recalls, an access state indicates whether the user accesses the recall commodity or not, and a binarization identification representation is adopted, for example, the recall commodity accessed by the user is represented as 1, and otherwise, the recall commodity is represented as 0.
In the data cleaning process, any access form of the user to the recalled goods, such as clicking a corresponding link, adding the recalled goods to a shopping cart, performing an ordering operation on the recalled goods, paying an order corresponding to the recalled goods, and the like, can be normalized into a single access fact, that is, only whether the user accesses the recalled goods in any form is marked, and the access times of the user are not concerned. Therefore, the related calculation amount and complexity are simplified, and the execution efficiency is improved.
Step S1220, determining parameter values corresponding to the evaluation indexes of each recall source according to the user behavior data;
in order to apply the formulas corresponding to the evaluation indexes conveniently, the parameter values required by the parameters of the formulas are determined from the cleaned user behavior data based on the formulas, and it can be known by examining the exemplary formulas in the present application that the commodities in the commodity data subsets of the recall sources are classified into two types, namely visited and unvisited, the visited commodity is taken as a positive sample of the recall source, and the unvisited commodity is taken as a negative sample of the recall source, so that a confusion matrix of each recall source can be obtained, and the corresponding parameter values can be counted according to the confusion matrix.
In an example, according to the user behavior data after data cleaning, the commodities visited by the user in the commodity data subsets of the recall sources are counted, and the quantity Click of the commodities visited can be obtainedSingleAnd the quantity of non-accessed commodities of each recall source is unClickSingleAnd the evaluation index can be obtained through statistics correspondingly, so that the parameter value corresponding to each evaluation index is determined.
Step S1230, calculating the evaluation index corresponding to each recall source by using the preset algorithm corresponding to each evaluation index and the corresponding parameter value thereof.
After determining the parameter values of each evaluation index, such as the number of accessed and unaccessed commodities, the evaluation index formulas referring to the parameter values can be applied to each evaluation index in the example of the application, so that each evaluation index is correspondingly calculated, and finally, the corresponding result value of each evaluation index is obtained.
According to the embodiment, the parameter values required by the evaluation indexes can be rapidly counted on the basis of the user behavior data, the evaluation indexes of the recall sources are determined according to the user behavior data, and the evaluation of the same standard can be implemented on the advantages and the disadvantages of the recall sources.
Based on any of the above embodiments, referring to fig. 3, in step S1300, respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source, the method includes the following steps:
step 1310, sorting all the recall sources according to each evaluation index to obtain a recall source sorted list corresponding to each evaluation index;
for each evaluation index, the recall sources may be sorted respectively, and one recall source sorted list is obtained accordingly, for example, five evaluation indexes in the above example, and five recall source sorted lists may be obtained accordingly. For ease of understanding, the recall sources are ranked here for each recall source ranking list by their respective rating index from high to low.
Step S1320, according to the numerical value of the evaluation index, a uniform sorting score sequence is applied, and corresponding sorting scores from high to low are set for the recall sources in each recall source sorting list;
a rank score sequence is preset, for example, if N recall sources exist, where N is a natural number greater than 2, then the rank score sequence may be preset as follows:
[1,2,……N]
then, for each recall source ranking list, the ranking scores in the ranking score sequence are adapted according to the high and low evaluation indexes from high to low, for example, the recall source matched ranking score N with the highest evaluation index, the recall source matched ranking score 1 with the lowest evaluation index, and the like.
And step S1330, summing the ranking scores of the recall source ranking list corresponding to the evaluation indexes by matching preset weights, and obtaining the index score corresponding to each recall source.
And each recall source realizes the conversion of the dimensions of different evaluation indexes to the same numerical value space for representing the ranking scores through the ranking scores determined by the recall sources corresponding to the evaluation indexes, so that each recall source can correspond to the preset weight of the different evaluation indexes, the ranking scores of the different evaluation indexes corresponding to the recall sources are weighted and summed, and the summed result can be used as the index score corresponding to each recall source.
For example, the following formula is used:
Scoresource＝w1*Rankprecision+w2*Rankrecall+w3*Rankf+w4*Rankauc+w5*Rankgauc
wherein, Rankprecision、Rankrecall、Rankf、Rankauc、RankgaucAnd the ranking scores are respectively corresponding to the evaluation indexes such as the accuracy, the recall ratio, the harmonic score, the area under the curve, the area under the user curve and the like, and the source is a recall source. w is a predetermined weight, which can be predetermined by one skilled in the art. In an exemplary weight configuration scheme, the area under the curve and the area under the curve of the user can maintain the same highest weight, the recall rate and the harmonic score can maintain the same lowest weight, and the accuracy rate can be a compromise weight between the two.
According to the embodiment, the ranking score is used as the intermediate dimension, the equivalent analysis corresponding to each evaluation index is realized, the effect of each evaluation index is conveniently subjected to standardized adjustment through the preset weight, and therefore the finally obtained index score is adjusted, and objective and effective evaluation results are rapidly obtained.
Optionally, before the step of obtaining the plurality of subsets of commodity data recalled by the preset recall sources, the method includes the following steps:
and responding to the commodity matching instruction, calling a plurality of recall sources, and matching out a corresponding commodity data subset for the target commodity specified by the commodity matching instruction based on each recall source, wherein the commodities in the commodity data subset are similar to the target commodity in composition characteristics.
In an exemplary application scenario, a user triggers a commodity matching instruction at an online store of an independent site of an e-commerce platform and sends the commodity matching instruction to a server, wherein the commodity matching instruction comprises a specified target commodity so as to implement similar commodity matching, for this purpose, the server enables a corresponding recall source to recall the commodity in response to the commodity matching instruction, and accordingly, a commodity data subset obtained by retrieving the similar commodity of each recall source according to the target commodity is obtained.
In one embodiment, when each recall source searches for similar commodities for a target commodity, a data distance between deep semantic information pre-extracted from commodities in a commodity data pool corresponding to the recall source is calculated based on deep semantic information pre-extracted from commodity information of the commodities, so as to determine similarity between the deep semantic information and the deep semantic information, and then, according to a preset threshold, commodities with similarity higher than the preset threshold are searched from the commodity data pool to form a corresponding commodity data subset as a recall result. In the commodity data subset, the similarity corresponding to each commodity may be stored in association, and the similarity is used as the matching degree between the corresponding commodity and the target commodity.
According to the embodiment, the method and the device for the commodity recommendation can respond to the user instruction through the pre-step, realize user interaction, execute corresponding recall according to the instruction of the terminal device, and then determine the commodity recommendation set matched with the target commodity on the basis.
Optionally, referring to fig. 4, after the step of merging the commodity data subsets into a commodity recommendation set, the method includes the following steps:
s1500, inversely ordering the commodity recommendation set according to the actual scores of the commodities in the commodity recommendation set;
as described above, in the product recommendation set formed by merging the product data subsets of the recall sources, each product obtains its corresponding actual score, and the actual score can effectively and uniformly represent the information contribution value of each product by dimension conversion and weight matching on the basis of the evaluation index of each recall source, so that each product in the product recommendation set can be directly sorted inversely according to the actual score to obtain a sorting result from large to small.
Step S1600, obtaining a preset plurality of commodities which are ranked at the front from the commodity recommendation set to form a commodity recommendation list;
generally, for some specific application scenarios, such as similar matching of commodities, placement of advertised commodities, and the like, after a commodity recommendation set is obtained based on recall of target commodities, the quantity of commodities in the commodity recommendation set is large, for example, hundreds or thousands, but in practical application, only a part of commodities with high information contribution value need to be displayed, so that according to a preset quantity, the commodities in the sorted commodity recommendation set can be truncated, a preset quantity of commodities in the front-ranked part of commodities is reserved, and other commodities behind the commodities are removed, thereby obtaining a commodity recommendation list.
And S1700, pushing the commodity recommendation list to target terminal equipment.
And finally, pushing the commodity recommendation list to a terminal device driving commodity recall, such as a terminal device of a user providing the target commodity, according to an actual scene, so that the commodity recommendation list can be analyzed and displayed on a graphical user interface of the terminal device.
According to the embodiments, the commodity recommendation list is quickly obtained with a low calculation amount, the requirements such as commodity similarity matching, advertisement commodity putting and the like can be met, the method and the system are particularly suitable for being deployed in online shops of an e-commerce platform realized on the basis of independent sites, the deployment cost is low, and the obtained recall result is accurate and efficient.
Referring to fig. 5, a merchandise recall optimization apparatus adapted to one of the objectives of the present application is provided, which is a functional implementation of the merchandise recall optimization method of the present application, and the apparatus includes a data acquisition module 1100, an index determination module 1200, an index summarization module 1300, and a data merging module 1400, wherein: the data acquiring module 1100 is configured to acquire a plurality of preset commodity data subsets recalled by a plurality of recall sources, where each subset includes a recalled commodity and a matching degree representing the recalled commodity; the index determining module 1200 is configured to determine a plurality of evaluation indexes of each recall source according to user behavior data corresponding to the commodities in each commodity data subset; the index summarizing module 1300 is configured to summarize and determine an index score of each recall source according to each evaluation index corresponding to each recall source; the data merging module 1400 is configured to merge the commodity data subsets into a commodity recommendation set, where the actual score of each commodity is ranked according to the actual score of each commodity, and the actual score of each commodity is a product of a matching degree of the commodity and an index score of a recall source of the commodity.
Optionally, the index determining module 1200 includes: the behavior acquisition unit is used for acquiring user behavior data corresponding to the commodities in the commodity data subset corresponding to each recall source; the parameter determining unit is used for determining parameter values corresponding to the evaluation indexes of each recall source according to the user behavior data; and the index calculation unit is used for applying the preset algorithm corresponding to each evaluation index and the corresponding parameter value thereof to calculate the evaluation index corresponding to each recall source.
Optionally, the index summarizing module 1300 includes: the sorting processing unit is used for sorting the recall sources according to each evaluation index to obtain a recall source sorting list corresponding to each evaluation index; the sorting assignment unit is used for applying a uniform sorting score sequence according to the numerical value of the evaluation index and setting a corresponding sorting score from high to low for the recall sources in each recall source sorting list; and the summation unit is used for matching the preset weights with the sorting scores of the recall source sorted list corresponding to each evaluation index for summation to obtain the index score corresponding to each recall source.
Optionally, prior to the data acquiring module 1100, the data acquiring module includes: and the recall execution module is used for responding to the commodity matching instruction, calling a plurality of recall sources, and matching out a corresponding commodity data subset for the target commodity specified by the commodity matching instruction based on each recall source, wherein the commodities in the commodity data subset have similar composition characteristics with the target commodity.
Optionally, the data merging module 1400 includes: the merging and sequencing module is used for performing reverse sequencing on the commodity recommendation set according to the actual scores of the commodities in the commodity recommendation set; the list generation module is used for acquiring a preset plurality of commodities which are ranked at the front from the commodity recommendation set to form a commodity recommendation list; and the list pushing module is used for pushing the commodity recommendation list to target terminal equipment.
Optionally, the evaluation index includes any one or more of the following items: recall rate, accuracy rate, harmonic score, area under the curve, area under the user curve, the harmonic score being determined according to the recall rate and accuracy rate.
In order to solve the technical problem, an embodiment of the present application further provides a computer device. As shown in fig. 6, the internal structure of the computer device is schematically illustrated. The computer device includes a processor, a computer-readable storage medium, a memory, and a network interface connected by a system bus. The computer readable storage medium of the computer device stores an operating system, a database and computer readable instructions, the database can store control information sequences, and when the computer readable instructions are executed by a processor, the processor can realize a commodity search category identification method. The processor of the computer device is used for providing calculation and control capability and supporting the operation of the whole computer device. The memory of the computer device may have stored therein computer readable instructions that, when executed by the processor, may cause the processor to perform the article recall optimization method of the present application. The network interface of the computer device is used for connecting and communicating with the terminal. Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In this embodiment, the processor is configured to execute specific functions of each module and its sub-module in fig. 5, and the memory stores program codes and various data required for executing the modules or sub-modules. The network interface is used for data transmission to and from a user terminal or a server. The memory in the present embodiment stores program codes and data necessary for executing all modules/sub-modules in the product recall optimization apparatus of the present application, and the server can call the program codes and data of the server to execute the functions of all sub-modules.
The present application further provides a storage medium having stored thereon computer-readable instructions which, when executed by one or more processors, cause the one or more processors to perform the steps of the method for merchandise recall optimization of any of the embodiments of the present application.
The present application also provides a computer program product comprising computer programs/instructions which, when executed by one or more processors, implement the steps of the method as described in any of the embodiments of the present application.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments of the present application may be implemented by hardware related to instructions of a computer program, where the computer program may be stored in a computer-readable storage medium, and when executed, the computer program may include the processes of the embodiments of the methods as described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), or other computer readable storage medium, or a Random Access Memory (RAM).
To sum up, the method and the device for recommending the commodities have the advantages that the fine granularity quantification of the information contribution degrees of the commodities obtained by different recall sources is achieved, so that a more accurate final recommendation result can be determined according to the actual scores of the commodities, a complex mathematical model is not needed in the implementation process, the method and the device can be achieved with lower system overhead, the deployment cost is lower, and the method and the device are particularly suitable for independent sites in an e-commerce platform.
Those of skill in the art will appreciate that the various operations, methods, steps in the processes, acts, or solutions discussed in this application can be interchanged, modified, combined, or eliminated. Further, other steps, measures, or schemes in various operations, methods, or flows that have been discussed in this application can be alternated, altered, rearranged, broken down, combined, or deleted. Further, steps, measures, schemes in the prior art having various operations, methods, procedures disclosed in the present application may also be alternated, modified, rearranged, decomposed, combined, or deleted.
The foregoing is only a few embodiments of the present application and it should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present application, and that these improvements and modifications should also be considered as the protection scope of the present application.
Claims (10)
1. A method for merchandise recall optimization, comprising the steps of:
obtaining a plurality of preset commodity data subsets recalled correspondingly by a plurality of recall sources, wherein each subset comprises recalled commodities and a matching degree representing the recalled commodities;
determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset;
respectively summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source;
and merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity.
2. The merchandise recall optimization method of claim 1 wherein determining a plurality of evaluation metrics for each recall source based on user behavior data corresponding to the merchandise in each subset of merchandise data comprises the steps of:
acquiring user behavior data corresponding to commodities in a commodity data subset corresponding to each recall source;
determining parameter values corresponding to all evaluation indexes of each recall source according to the user behavior data;
and calculating the evaluation indexes corresponding to the recall sources by applying the preset algorithm corresponding to each evaluation index and the corresponding parameter values thereof.
3. The merchandise recall optimization method of claim 1, wherein the step of determining the index score of each recall source is performed by summarizing the respective evaluation indexes of each recall source, respectively, and comprises the steps of:
sorting all the recall sources according to each evaluation index to obtain a recall source sorting list corresponding to each evaluation index;
according to the numerical value of the evaluation index, a uniform ranking score sequence is applied, and corresponding ranking scores from high to low are set for the recall sources in each recall source ranking list;
and matching the preset weights with the ranking scores of the recall source ranking list corresponding to each evaluation index, and summing to obtain the index score corresponding to each recall source.
4. The merchandise recall optimization method according to any one of claims 1 to 3, wherein the step of obtaining a plurality of merchandise data subsets corresponding to the recalls of the plurality of preset recall sources is preceded by the steps of:
and responding to the commodity matching instruction, calling a plurality of recall sources, and matching out a corresponding commodity data subset for the target commodity specified by the commodity matching instruction based on each recall source, wherein the commodities in the commodity data subset are similar to the target commodity in composition characteristics.
5. The merchandise recall optimization method of any one of claims 1 to 3 wherein the step of merging the subset of merchandise data into a recommendation set of merchandise is followed by the steps of:
performing reverse ordering on the commodity recommendation set according to the actual scores of the commodities in the commodity recommendation set;
obtaining a preset plurality of commodities which are ranked at the front from the commodity recommendation set to form a commodity recommendation list;
and pushing the commodity recommendation list to target terminal equipment.
6. The product recall optimization method of any one of claims 1 to 3 wherein the evaluation index comprises any one or more of: the system comprises a recall rate, an accuracy rate, a reconciliation score, an area under a curve and an area under a user curve, wherein the reconciliation score is determined according to the recall rate and the accuracy rate.
7. An article recall optimization apparatus, comprising:
the data acquisition module is used for acquiring a plurality of preset commodity data subsets recalled by corresponding recall sources, and each subset comprises recalled commodities and a matching degree representing the recalled commodities;
the index determining module is used for determining a plurality of evaluation indexes of each recall source according to user behavior data corresponding to commodities in each commodity data subset;
the index summarizing module is used for summarizing and determining the index score of each recall source according to each evaluation index corresponding to each recall source;
and the data merging module is used for merging the commodity data subsets into a commodity recommendation set, wherein the commodity recommendation set is sorted according to the actual score of each commodity, and the actual score of each commodity is the product of the matching degree of the commodity and the index score of the recall source of the commodity.
8. A computer device comprising a central processor and a memory, characterized in that the central processor is adapted to invoke execution of a computer program stored in the memory to perform the steps of the method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that it stores, in the form of computer-readable instructions, a computer program implemented according to the method of any one of claims 1 to 6, which, when invoked by a computer, performs the steps comprised by the corresponding method.
10. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210555827.3A CN114782062A (en) | 2022-05-20 | 2022-05-20 | Commodity recall optimization method and device, equipment, medium and product thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210555827.3A CN114782062A (en) | 2022-05-20 | 2022-05-20 | Commodity recall optimization method and device, equipment, medium and product thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114782062A true CN114782062A (en) | 2022-07-22 |
Family
ID=82408722
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210555827.3A Pending CN114782062A (en) | 2022-05-20 | 2022-05-20 | Commodity recall optimization method and device, equipment, medium and product thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114782062A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117455579A (en) * | 2023-12-25 | 2024-01-26 | 深圳市智岩科技有限公司 | Commodity recommendation intervention method, commodity recommendation intervention device, medium and equipment |
-
2022
- 2022-05-20 CN CN202210555827.3A patent/CN114782062A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117455579A (en) * | 2023-12-25 | 2024-01-26 | 深圳市智岩科技有限公司 | Commodity recommendation intervention method, commodity recommendation intervention device, medium and equipment |
| CN117455579B (en) * | 2023-12-25 | 2024-04-09 | 深圳市智岩科技有限公司 | Commodity recommendation intervention method, commodity recommendation intervention device, medium and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108230007B (en) | User intention identification method and device, electronic equipment and storage medium | |
| US20160092912A1 (en) | Automated optimization of an advertisement placement engine | |
| CN110110233B (en) | Information processing method, device, medium and computing equipment | |
| CN109255000B (en) | Dimension management method and device for label data | |
| CN111898247B (en) | Landslide displacement prediction method, landslide displacement prediction equipment and storage medium | |
| US20210233147A1 (en) | Recommendation engine based on optimized combination of recommendation algorithms | |
| CN113627995A (en) | Commodity recommendation list updating method and device, equipment, medium and product thereof | |
| US8095482B2 (en) | System and method to learn and deploy an optimal user experience in an online system | |
| KR20230095796A (en) | Joint personalized search and recommendation with hypergraph convolutional networks | |
| CN113327151A (en) | Commodity object recommendation method and device, computer equipment and storage medium | |
| CN115423555A (en) | Commodity recommendation method and device, electronic equipment and storage medium | |
| CN114782062A (en) | Commodity recall optimization method and device, equipment, medium and product thereof | |
| CN114781717A (en) | Network point equipment recommendation method, device, equipment and storage medium | |
| CN113742578A (en) | Data recommendation method and device, electronic equipment and storage medium | |
| CN112464106A (en) | Object recommendation method and device | |
| CN112348300A (en) | Method and device for pushing information | |
| US9779142B2 (en) | Methods and systems to estimate query responses based on data set sketches | |
| CN112967100B (en) | Similar crowd expansion method, device, computing equipment and medium | |
| CN114169965A (en) | Commodity similarity matching method and device, equipment, medium and product thereof | |
| CN115293818A (en) | Advertisement putting and selecting method and device, equipment and medium thereof | |
| CN110852078A (en) | Method and device for generating title | |
| CN114971766A (en) | Commodity recommendation method and device, equipment, medium and product thereof | |
| EP4134898A1 (en) | Method and apparatus for pushing information | |
| CN109325198B (en) | Resource display method and device and storage medium | |
| CN114638316A (en) | Data clustering method, device and equipment |
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 |