An object behavior-based item recommendation method, device, equipment and medium

Abstract

This application belongs to the field of big data processing technology and provides a method, apparatus, device, and medium for item recommendation based on object behavior. The method acquires tracking data from multiple objects, generates a first item set based on the information in the tracking data, obtains candidate recommended items based on the similarity between items in the first item set, and then calculates a recommendation score by acquiring the preference weight coefficient, behavior weight coefficient, time decay coefficient, and second similarity of each candidate recommended item. The recommended items are then ranked according to the recommendation scores to obtain target recommended items and recommended to the target object. By incorporating preference weight coefficients, behavior weight coefficients, and time decay coefficients—which characterize the target object's preference for items—the recommendation results are optimized, improving the degree of personalized item recommendation.

Description

Technical Field

[0001] This application relates to the field of big data processing technology, and in particular to a method, apparatus, device and medium for recommending items based on object behavior. Background Technology

[0002] Currently, traditional item recommendation methods can calculate the similarity between items and recommend items with high similarity to the target object. The core of this method is that the target object who likes item A also likes item B. However, this method does not consider the object's access behavior and the degree of preference for the item, and cannot make personalized recommendations for the target object. Summary of the Invention

[0003] The main objective of this application is to propose a method, apparatus, electronic device, and medium for recommending items based on object behavior, aiming to improve the degree of personalized recommendations.

[0004] To achieve the above objectives, a first aspect of this application proposes an item recommendation method based on object behavior, the method comprising the following steps:

[0005] Collect tracking data of multiple objects. The tracking data includes items accessed by the objects within a preset time period and access information corresponding to the items. The access information includes access type and access time.

[0006] The items visited by the target object are obtained from the data points, and a first set of items is generated, wherein the target object is one of the plurality of objects;

[0007] Based on the similarity between the items in the first item set, N candidate recommended items are determined, where N is a preset value greater than 1;

[0008] For each candidate recommended item, the preference weight coefficient of the candidate recommended item for the target object is determined according to the first preset algorithm;

[0009] Based on the access type of the candidate recommended items to the target object, determine the behavioral weight coefficient of the target object for the candidate recommended items;

[0010] Based on the access time of the candidate recommended items to the target object, determine the time decay coefficient of the candidate recommended items for the target object;

[0011] For each candidate recommended item, a recommendation score is determined based on the candidate recommended item's preference weight coefficient, behavior weight coefficient, time decay coefficient, and second similarity to the target object.

[0012] determine a target recommended item according to a recommendation score of each of the candidate recommended items, and recommend the target recommended item to the target object.

[0013] In some possible embodiments, the determining the N candidate recommended items according to the similarity between the items in the first item set comprises:

[0014] For any two items in the first item set, a first similarity between the two items is determined according to the number of times each of the two items is accessed by the plurality of objects and the number of common access objects of the two items.

[0015] For each of the items in the first item set, a second similarity of the item is obtained according to a sum of the first similarities between the item and other items in the first item set.

[0016] The items in the first item set are sorted in descending order of the second similarity to obtain an item sequence.

[0017] The first N items in the item sequence are determined as the candidate recommended items, where N is a preset value greater than 1.

[0018] In some possible embodiments, the determining, for each of the candidate recommended items, the preference weight coefficient of the candidate recommended item for the target object according to a first preset algorithm comprises:

[0019] The preference weight coefficient of the candidate recommended item for the target object is calculated according to the first preset algorithm, the formula of the first preset algorithm being as follows:

[0020]

[0021] wherein, T represents the preference weight coefficient.

[0022] a represents the number of times the target object accesses the currently traversed candidate recommended item.

[0023] b represents the number of times the target object accesses all the items in the first item set.

[0024] c represents the number of times the plurality of objects access all the candidate recommended items.

[0025] d represents the number of times the plurality of objects access the currently traversed candidate recommended item.

[0026] e is a natural constant.

[0027] In some possible embodiments, the determining the time decay coefficient of the candidate recommended item for the target object according to the access time of the candidate recommended item for the target object comprises:

[0028] determining a time interval parameter according to the access time of the candidate recommended item for the target object;

[0029] determining the time decay coefficient according to a second preset algorithm and the time interval parameter, a formula of the second preset algorithm being as follows:

[0030] F2=F1×e (-α×t)

[0031] wherein, the F2 represents the interest degree corresponding to the time interval parameter;

[0032] the F1 is a preset initial interest degree;

[0033] the a represents the time decay coefficient;

[0034] the t is the time interval parameter.

[0035] In some possible embodiments of the present application, a formula of the first similarity is as follows:

[0036]

[0037] wherein, the i represents an i-th item in the first item set, and the j represents a j-th item in the first item set;

[0038] the sim[i][j] represents the first similarity between the i-th item and the j-th item;

[0039] the cnt[i] represents the access times of the i-th item accessed by the plurality of objects;

[0040] the cnt[j] represents the access times of the j-th item accessed by the plurality of objects;

[0041] the mp[i][j] represents the number of common access objects of the i-th item and the j-th item;

[0042] the g represents an active penalty coefficient.

[0043] In some possible embodiments, a formula of the active penalty coefficient is as follows:

[0044]

[0045] wherein, the |itemlist| represents the number of items included in the first item set.

[0046] In some possible embodiments, the calculation formula of the recommendation score of the candidate recommended item is as follows:

[0047] score=T*G*alpha*beta

[0048] wherein the score represents the recommendation score of the candidate recommended item;

[0049] the T represents the preference weight coefficient;

[0050] the G represents the behavior weight coefficient;

[0051] the alpha represents the time decay coefficient;

[0052] the beta represents the second similarity of the candidate recommended item.

[0053] To achieve the above object, a second aspect of the embodiment of the present application provides a device for recommending items based on object behaviors, which comprises:

[0054] a buried point data collection module, configured to collect buried point data of a plurality of objects, the buried point data comprising items accessed by the objects within a preset time period and access information corresponding to the items, the access information comprising an access type and an access time;

[0055] a first item set generation module, configured to obtain items accessed by a target object from the buried point data and generate a first item set, the target object being one of the plurality of objects;

[0056] a candidate recommended item acquisition module, configured to determine N candidate recommended items according to similarities between the items in the first item set, the N being a preset value greater than 1;

[0057] a preference weight coefficient acquisition module, configured to determine, for each candidate recommended item, a preference weight coefficient of the candidate recommended item for a target object according to a first preset algorithm;

[0058] a behavior weight coefficient acquisition module, configured to determine a behavior weight coefficient of the target object for the candidate recommended item according to an access type of the candidate recommended item for the target object;

[0059] a time decay coefficient acquisition module, configured to determine a time decay coefficient of the candidate recommended item for the target object according to an access time of each candidate recommended item for the target object;

[0060] The recommendation score acquisition module is used to determine the recommendation score of each candidate recommendation item based on the preference weight coefficient, behavior weight coefficient, time decay coefficient and second similarity of the candidate recommendation item to the target object.

[0061] The item recommendation module is used to determine the target recommended item based on the recommendation scores of each of the candidate recommended items, and to recommend the target recommended item to the target object.

[0062] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0063] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0064] This application proposes a method, apparatus, electronic device, and storage medium for item recommendation based on object behavior. It acquires tracking data from multiple objects, generates a first item set based on the information in the tracking data, obtains candidate recommended items based on the similarity between items in the first item set, and then calculates a recommendation score by acquiring the preference weight coefficient, behavior weight coefficient, time decay coefficient, and a second similarity of each candidate recommended item. The recommended items are then ranked according to the recommendation scores to obtain target recommended items and recommended to the target object. By incorporating preference weight coefficients, behavior weight coefficients, and time decay coefficients—which characterize the target object's preference for items—the recommendation results are optimized, improving the degree of personalized item recommendations. Attached Figure Description

[0065] Figure 1 This is a flowchart illustrating an item recommendation method based on object behavior provided in an embodiment of this application;

[0066] Figure 2 yes Figure 1 A schematic diagram of the sub-steps in step S103;

[0067] Figure 3 yes Figure 1 A schematic diagram of the sub-steps in step S106;

[0068] Figure 4 yes Figure 1 A schematic diagram of the sub-steps in step S108;

[0069] Figure 5is a flow diagram of a method for recommending an item based on an object behavior according to another embodiment of the present application;

[0070] Figure 6 is a flow diagram of a method for recommending an item based on an object behavior according to an embodiment of the present application;

[0071] Figure 7 is a structural diagram of an apparatus for recommending an item based on an object behavior according to an embodiment of the present application;

[0072] Figure 8 is a structural diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0073] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application.

[0074] It should be noted that although the functional modules are divided in the apparatus diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a manner different from the module division in the apparatus or the order in the flowchart. The terms "first", "second", etc. in the specification and claims and the above-described drawings are used to distinguish similar objects and do not necessarily describe a specific order or sequence.

[0075] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0076] First, the terms involved in the present application are analyzed:

[0077] Big data: refers to data that is larger than general collected data, which is a data set that is so large that it greatly exceeds the capabilities of traditional database software tools in terms of acquisition, storage, management and analysis. Big data requires special technology to effectively process large amounts of data over time. The technologies suitable for big data include massively parallel processing (MPP) databases, data mining, distributed file systems, distributed databases, cloud computing platforms, the Internet and scalable storage systems.

[0078] The embodiments of the present application provide a method, apparatus, device and medium for recommending an item based on an object behavior. The embodiments are specifically described as follows: first, a method for recommending an item based on an object behavior according to an embodiment of the present application is described.

[0079] The embodiments of the present application can acquire and process related data based on big data technology. The big data technology is a theory, method, technology and application system for collecting a large amount of data, analyzing the data and obtaining different inferences or analysis conclusions based on the data analysis results by using a digital computer or a machine controlled by a digital computer.

[0080] The object behavior-based item recommendation method provided by the embodiments of the present application relates to the technical field of big data processing. The object behavior-based item recommendation method provided by the embodiments of the present application can be applied to a terminal, can be applied to a server end, and can also be software running in the terminal or the server end. In some embodiments, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc.; the server end can be configured as a stand-alone physical server, can be configured as a server cluster or a distributed system composed of multiple physical servers, can also be configured as a cloud server providing basic cloud computing services such as cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and big data and artificial intelligence platform; and the software can be an application program for implementing an object behavior-based item recommendation method, but is not limited to the above forms.

[0081] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0082] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards of the relevant countries and regions. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirects to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data for the proper functioning of the embodiments of this application obtained.

[0083] Currently, traditional item recommendation methods calculate the similarity between items and recommend highly similar items to the target user. The core of this method is that a target user who likes item A also likes item B. However, this method does not consider the user's browsing behavior and the degree of preference for items, thus failing to provide personalized recommendations. Therefore, this application proposes an item recommendation method, apparatus, electronic device, and medium based on user behavior, aiming to improve the degree of personalized recommendations.

[0084] Please see Figure 1 , Figure 1 This is a flowchart illustrating an item recommendation method based on object behavior provided in an embodiment of this application. Figure 1 The method may include, but is not limited to, steps S101 to S111.

[0085] Step S101: Collect data from the tracking points of multiple objects.

[0086] It should be understood that the tracking data here includes the items that each object has accessed within a preset time period, and the access information corresponding to the items. The access information includes the access type and access time. The preset time period is Q days, where Q is greater than 1.

[0087] It should be understood that the access type here can be browsing, favorites, adding to cart, purchasing, etc. Those skilled in the art can set up tracking points for different access behaviors to obtain access type data according to the actual situation, and this application does not limit this.

[0088] It should be understood that the collected data is stored in a data tracking table. Those skilled in the art can obtain different data from the contents of any table based on different table entries, according to the actual situation. This application does not limit this.

[0089] Specifically, based on the object tag corresponding to each object, the data collection is completed by extracting the trigger data corresponding to each object from the collected event tracking data.

[0090] Step S102: Obtain the items accessed by the target object from the buried point data, and generate a first item set.

[0091] It should be understood that the target object here is one of the multiple objects whose buried point data is collected in step S101.

[0092] It should be understood that before step S102 is implemented, the collected multiple object buried point data is preprocessed, and is divided according to the collection factors of the buried point data. Those skilled in the art can divide the buried point data according to different collection factors such as specific time, access type, customer type, etc. according to the actual situation to complete the preprocessing of the data. The present application does not limit this. Illustratively, the object's item access behavior can be divided into each day according to the access time in the buried point data, and the buried point data divided into each day is further divided according to the access type.

[0093] Specifically, according to the preprocessed buried point data, the items accessed by the target object are obtained, and the accessed items are generated into a first item set.

[0094] It should be understood that the set here is a way of data set, which can be a data set, a data class, a data table, etc. Those skilled in the art can select the corresponding data set way to generate the first item set according to the actual situation. The present application does not limit this.

[0095] Step S103: Determine N candidate recommended items according to the similarity between the items in the first item set.

[0096] Please refer to Figure 2 , Figure 2 Figure 1 The sub-step schematic diagram of step S103 in the method. In some possible embodiments of the present application, step S103 includes but is not limited to the following sub-steps:

[0097] Step S301: For any two items in the first item set, determine the first similarity between the two items according to the access times of each item in the two items accessed by multiple objects and the number of common access objects of the two items.

[0098] It should be understood that the any two items here refer to selecting one item in the first item set and selecting any one item from the other items except this item to form the above-mentioned any two items.

[0099] Specifically, based on the access times of each of the two items accessed by the plurality of objects and the common access object number of the two items, the similarity between the two items is determined according to a first similarity calculation formula. The first similarity calculation formula is as follows:

[0100]

[0101] wherein, i represents the i-th item in the first item set, that is, one selected from the first item set, j represents the j-th item in the first item set, that is, any one selected from the other items except the i-th item;

[0102] sim[i][j] represents the first similarity between the i-th item and the j-th item;

[0103] cnt[i] represents the access times of the i-th item accessed by the plurality of objects;

[0104] cnt[j] represents the access times of the j-th item accessed by the plurality of objects;

[0105] mp[i][j] represents the common access object number of the i-th item and the j-th item;

[0106] γ represents an active penalty coefficient, and the active penalty coefficient is calculated according to the following formula:

[0107]

[0108] wherein, |itemlist| represents the number of items included in the first item set.

[0109] By increasing the active penalty coefficient, the similarity misjudgment caused by the frequent access of active objects to items is weakened, so as to reduce the recommendation error of the items to non-active objects.

[0110] Step S302: For each of the items in the first item set, the second similarity of the item is obtained according to the sum of the first similarities between the item and other items in the first item set.

[0111] Specifically, in step S301, the first similarity of each of the items in the first item set with respect to each of the other items is obtained, that is, the similarity between any two of the items, but not the similarity of each of the items to the entire first item set, therefore, the sum of the values of all the first similarities is obtained by adding all the first similarities, and the second similarity is obtained, which is used as the similarity of each of the items to the first item set.

[0112] Step S303: The items in the first item set are sorted in descending order of the second similarity, and an item sequence is obtained.

[0113] Specifically, since the second similarity obtained in step S302 represents the similarity of each item to the entire first item set, the higher the similarity, the more similar items there are to the entire first item set. The items are sorted according to the value of the second similarity, and the items with a high probability of being of the same type are selected from the first item set and determined as an item sequence.

[0114] Step S304: Determine the first N items in the item sequence as candidate recommended items.

[0115] It should be understood that N here is a preset value greater than 1. Those skilled in the art can set the specific value of N according to the actual situation, and this application does not limit it in this regard.

[0116] In some possible embodiments of this application, N can be obtained by setting a preset coefficient for N based on the number of items in the item sequence obtained in step S303, so that N is greater than 1. For example, if there are two items in the sequence, a percentage coefficient of 50% is set, and the first-ranked item is selected as the candidate recommended item; for another example, if there are four items in the sequence, a percentage coefficient of 25% is set, and the first-ranked item is selected as the candidate recommended item, etc. Those skilled in the art can set the coefficient for N according to the actual situation, and this application does not limit it.

[0117] In some possible embodiments of this application, the value of N is obtained in segments based on the number of items in the item sequence obtained in step S303, so that N is greater than 1. For example, if there are two items in the item sequence, the first ranked item is taken as a candidate recommended item; if the number of items in the item sequence is greater than 3 and less than 5, the first and second ranked items are taken as candidate recommended items, etc. Those skilled in the art can set the segmented N according to the actual situation, and this application does not limit it.

[0118] Step S104: For each candidate recommended item, determine the preference weight coefficient of the candidate recommended item for the target object according to the first preset algorithm.

[0119] It should be understood that, for each candidate recommended item, the preference weight coefficient of the candidate recommended item for the target object is determined by traversing the candidate recommended item and calculating the tracking data of multiple objects collected in step S101 based on the first preset algorithm.

[0120] It should be understood that the formula for the first preset algorithm here is as follows:

[0121]

[0122] wherein, in the formula, T represents the preference weight coefficient;

[0123] a represents the number of visits of the target object to the currently traversed candidate recommended item;

[0124] b represents the number of visits of the target object to all items in the first item set;

[0125] c represents the number of visits of the plurality of objects to all candidate recommended items;

[0126] d represents the number of visits of the plurality of objects to the currently traversed candidate recommended item.

[0127] It should be understood that the preference weight coefficient here is a representation of the preference of the target object for a recommended item at the level of the number of visits. The more times the target object visits the recommended item and the fewer times the target object visits other items, the more the target object prefers the recommended item, which is represented in the formula as T = a / d. When the target object prefers the recommended item, the calculated result value obtained by the formula will be larger.

[0128] Step S105: determining a behavior weight coefficient of the target object for the candidate recommended item according to the visit type of the candidate recommended item for the target object.

[0129] It should be understood that the candidate recommended item here is the currently traversed candidate recommended item in step S104. The behavior weight coefficient here is a representation of the preference of the target object for a recommended item at the level of the visit type. Different visit types represent different preferences of the target object for the recommended item. For example, a target object that purchases a recommended item has a higher preference for the recommended item than a target object that collects the recommended item.

[0130] It should be understood that the behavior weight coefficient here is a preset value. A preset value is set for each visit type. Those skilled in the art can set the behavior weight coefficient of each visit type according to actual conditions, and the present application does not limit this.

[0131] In some possible embodiments of the present application, a preset value is set for each visit type as the behavior weight coefficient. The behavior weight coefficient of each visit type is universal for each recommended item. For example, the behavior weight coefficient of purchasing item A is 0.8, the behavior weight coefficient of purchasing item B is 0.8, and the behavior weight coefficient of the visit type of “purchasing” is 0.8.

[0132] In some possible embodiments of this application, each access type is set with multiple different preset values ​​as behavior bias coefficients. For different products, the same access type has different behavior bias coefficients. For example, when purchasing item A, the behavior bias coefficient is 0.8, and when purchasing item B, the behavior bias coefficient is 0.9. Under the same access type of "purchase", different items result in different behavior bias coefficients.

[0133] Step S106: Determine the time decay coefficient of the candidate recommended item for the target object based on the access time of the candidate recommended item for the target object.

[0134] It should be understood that the candidate recommended item here is the same as the candidate recommended item in step S105, which is the candidate recommended item currently being traversed in step S104.

[0135] Specifically, the time interval between the target object's access to the candidate recommended items is determined based on the access time of the candidate recommended items to the target object, and the time decay coefficient of the candidate recommended items to the target object with respect to the decay of interest is obtained based on the second preset algorithm, the time interval, and the degree of interest.

[0136] Please see Figure 3 , Figure 3 for Figure 1 A schematic diagram of the sub-steps of step S106 is shown. In some possible embodiments of this application, step S106 includes the following sub-steps:

[0137] Step S601: Determine the time interval parameter based on the access time of the candidate recommended items to the target object.

[0138] It should be understood that the time interval parameter here refers to the time interval between the target object accessing each candidate recommended item, which is the time difference between the specific time of collecting the tracking data of multiple objects in step S101 and the specific time of the target object accessing the candidate recommended item.

[0139] Step S602: Determine the time decay coefficient according to the second preset algorithm and time interval parameters.

[0140] It should be understood that the formula for the second preset algorithm here is as follows:

[0141] F2=F1×e (-α×t)

[0142] Where F2 represents the degree of interest corresponding to the time interval parameter;

[0143] F1 is the preset initial level of interest;

[0144] α represents the time decay coefficient;

[0145] t is a time interval parameter;

[0146] e is a natural constant.

[0147] It should be understood that F1 and F2 are preset values, and a person skilled in the art can set the specific values of F1 and F2 according to actual conditions, and the present application does not limit this.

[0148] In some embodiments of the present application, F2 is a preset value, and in any case, the interest degree of the target object to the candidate recommended item will only decrease to a fixed value.

[0149] In some embodiments of the present application, F2 is a dynamically obtained value, and a segmented value is set for F2, when the time interval coefficient meets a certain interval, the specific value of F2 corresponding to the interval is taken, and the interest degree of the target object to the candidate recommended item will only decrease to a fixed value.

[0150] Step S107: For each candidate recommended item, the recommendation score of the candidate recommended item is determined according to the preference weight coefficient, the behavior weight coefficient, the time decay coefficient of the candidate recommended item to the target object, and the second similarity of the recommended item.

[0151] It should be understood that the recommendation score is obtained by bringing the preference weight coefficient, the behavior weight coefficient, the time decay coefficient, and the second similarity of the recommended item into the calculation formula of the recommendation score, and the calculation formula of the recommendation score is as follows:

[0152] score=T×G×α×β

[0153] Wherein, score represents the recommendation score of the candidate recommended item;

[0154] T represents the preference weight coefficient;

[0155] G represents the behavior weight coefficient;

[0156] α represents the interest decay coefficient;

[0157] β represents the second similarity of the candidate recommended item.

[0158] Step S108: Determine the target recommended item according to the recommendation score of each candidate recommended item, and recommend the target recommended item to the target object.

[0159] Specifically, the candidate recommended items are sorted according to the recommendation score, and the first K in the sorting are determined as the target recommended items, and these target recommended items are recommended to the target object.

[0160] It should be noted that K here is a preset value, the value of K is greater than 0, and K here is less than or equal to N. Those skilled in the art can determine the specific value of K according to actual needs, and the present application does not limit this.

[0161] See Figure 4 , Figure 4 For Figure 1 The sub-step of step S108 is shown in the schematic diagram, and the sub-step of step S108 includes but is not limited to:

[0162] Step S801: Sort each candidate recommended item according to the recommended score from high to low to obtain a candidate recommended item sequence.

[0163] Step S802: Determine the first K candidate recommended items in the candidate recommended item sequence as target recommended items, and recommend the target recommended items to the target object.

[0164] In some possible embodiments of the present application, the value of K here can be obtained by setting a preset coefficient related to K according to the number of items in the candidate recommended item sequence obtained in step S801. For example, if there are two items in the sequence, a percentage coefficient of 50% is set, and the first ranked item is taken as the target recommended item. For another example, if there are four items in the sequence, a percentage coefficient of 25% is set, and the first ranked item is taken as the target recommended item. Those skilled in the art can set the coefficient related to N according to actual conditions, and the present application does not limit this.

[0165] In some possible embodiments of the present application, the value of K is obtained by segmenting according to the number of items in the candidate recommended item sequence obtained in step S801, so that K is greater than 1. For example, if there are two items in the sequence, the first ranked item is taken as the target recommended item. If the number of items in the sequence is greater than 3 and less than 5, the first and second ranked items are taken as the target recommended items. Those skilled in the art can set the segmented K according to actual conditions, and the present application does not limit this.

[0166] Steps S101 to S108, as illustrated in this embodiment, involve acquiring tracking data from multiple objects, generating a first item set based on the information in the tracking data, calculating the first similarity between any two items in the first item set, calculating the second similarity of each item to all items in the entire item set, sorting items according to the second similarity to obtain candidate recommended items, and then calculating a recommendation score by acquiring the preference weight coefficient, behavior weight coefficient, time decay coefficient, and the second similarity of each candidate recommended item. The recommendation scores are then sorted to obtain target recommended items for recommendation to the target object. By incorporating preference weight coefficients, behavior weight coefficients, and time decay coefficients—which characterize the target object's preference for items—the recommendation results are optimized, improving the degree of personalized item recommendations.

[0167] Please see Figure 5 , Figure 5 A flowchart illustrating an item recommendation method based on object behavior, provided for another embodiment of this application, includes, but is not limited to, step S109 after step S102:

[0168] Step S109: If there is only one item in the first item set, identify that item as the target recommended item and recommend it to the target object.

[0169] It should be noted that step S109 here is parallel to steps S103 to S108 above, and is a branch step. Implementing step S112 will not implement steps S103 to S108.

[0170] Please see Figure 6 , Figure 6 This is a schematic diagram illustrating the overall process of an item recommendation method based on object behavior, provided in an embodiment of this application. Figure 6 The specific steps of the item recommendation method based on object behavior provided in this application embodiment are the same as those described above, and will not be repeated here.

[0171] This application also provides an item recommendation device based on object behavior. Please refer to [link to relevant documentation]. Figure 7 , Figure 7 The schematic diagram of an object behavior-based item recommendation device 700 provided in this application embodiment can implement the above-mentioned object behavior-based item recommendation method. The device 700 includes:

[0172] The tracking data acquisition module 701 is used to collect tracking data of multiple objects. The tracking data includes items accessed by the objects within a preset time period and access information corresponding to the items. The access information includes access type and access time.

[0173] The first item set generation module 702 is used to obtain the items visited by the target object from the tracking data and generate the first item set, where the target object is one of multiple objects;

[0174] The candidate recommended item acquisition module 703 is used to determine N candidate recommended items based on the similarity between the items in the first item set, where N is a preset value greater than 1.

[0175] The preference weight coefficient acquisition module 704 is used to determine the preference weight coefficient of each candidate recommended item for the target object according to the first preset algorithm.

[0176] The behavior weight coefficient acquisition module 705 is used to determine the behavior weight coefficient of the target object for the candidate recommended items based on the access type of the candidate recommended items to the target object.

[0177] The time decay coefficient acquisition module 706 is used to determine the time decay coefficient of the candidate recommended item for the target object based on the access time of the recommended item for the target object.

[0178] The recommendation score acquisition module 707 is used to determine the recommendation score of each candidate recommendation item based on the preference weight coefficient, behavior weight coefficient, time decay coefficient and second similarity of the candidate recommendation item to the target object.

[0179] The item recommendation module 708 is used to determine the target recommended item based on the recommendation scores of each candidate recommended item, and recommend the target recommended item to the target object.

[0180] It should be noted that the specific implementation of the item recommendation device based on object behavior provided in this application is basically the same as the specific implementation of the item recommendation method based on object behavior described above, and will not be repeated here.

[0181] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned object behavior-based item recommendation method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0182] Please see Figure 8 , Figure 8 This application provides a schematic diagram of the structure of an electronic device 800, which includes:

[0183] The processor 801 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits, and is configured to execute related programs to implement the technical solutions provided by the embodiments of the present application.

[0184] The memory 802 can be implemented by a ROM (ReadOnly Memory), a static storage device, a dynamic storage device, or a RAM (Random Access Memory), and the like. The memory 802 can store an operating system and other application programs. When the technical solutions provided by the embodiments of the present application are implemented by software or firmware, the related program codes are stored in the memory 802 and are called and executed by the processor 801 to implement an object behavior-based item recommendation method according to an embodiment of the present application.

[0185] The input / output interface 803 is configured to implement information input and output.

[0186] The communication interface 804 is configured to implement the communication interaction between the device and other devices. The communication can be implemented by a wired manner (for example, a USB, a network cable, or the like) or a wireless manner (for example, a mobile network, WIFI, Bluetooth, or the like).

[0187] The bus 805 is configured to transmit information between various components (for example, the processor 801, the memory 802, the input / output interface 803, and the communication interface 804) of the device.

[0188] The processor 801, the memory 802, the input / output interface 803, and the communication interface 804 are connected to each other by the bus 805 to realize the communication connection between the devices.

[0189] The embodiments of the present application further provide a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the above-mentioned object behavior-based item recommendation method.

[0190] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory disposed remotely with respect to the processor, which can be connected to the processor through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0191] The embodiment of the application provides a kind of object behavior-based article recommendation method, device, equipment and medium, it obtains the burying point data of multiple objects, generates first article set according to the information in burying point data, obtains candidate recommended article according to the similarity between the articles in first article set, then obtains the preference weight coefficient, behavior weight coefficient, time attenuation coefficient of these candidate recommended article and the second similarity of each candidate recommended article, and calculates recommendation score, and target recommended article is obtained according to the order of recommendation score and is recommended to target object. By joining preference weight coefficient, behavior weight coefficient, time attenuation coefficient such as the degree of preference of target object to article, the recommendation result is optimized, and the degree of individualization of article is improved.

[0192] The embodiment described in the application embodiment is to more clearly illustrate the technical scheme of the application embodiment, and does not constitute a limitation on the technical scheme provided by the application embodiment. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical scheme provided by the application embodiment is also applicable to similar technical problems.

[0193] Those skilled in the art can understand that the technical scheme shown in the figure does not constitute a limitation on the application embodiment, and can include more or fewer steps than the figure, or combine certain steps, or different steps.

[0194] The device embodiments described above are only schematic, and the units described as separate components can or can not be physically separated, i.e., can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0195] Those skilled in the art can understand that all or some steps in the above disclosed method, the functions of the modules / units in the system and the device can be implemented as software, firmware, hardware and their appropriate combinations.

[0196] The terms "first", "second", "third", "fourth", and the like in the description and in the claims of this application, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of the terms so termed is interchangeable under appropriate circumstances such that the embodiments of the application described herein are, for example, capable of orderly or chronological mundane operation, reverse order operation, based on circuitry availability, based on stated preference or the like, and that "default" or other orderings are thus permissible. Further, the terms "comprise", "comprising", "include", "including", and the like, are specifically intended to be open-ended. That is, references to individual steps and the like do not suhstantially exclude the presence of two or more of a recited step or its integral sub-steps or additional steps whether or not readily ascertainable from the description or the like. Further, the words "a" or "an", as used herein in the disclosure and elsewhere, are used indiscriminately and are to be interpreted in the same way, i.e. as meaning "one or more".

[0197] It should be understood that, in this application, "at least one" means one or more, "multiple" means two or more. "And / or", used to describe the relationship between associated objects, means that there can be three relationships, for example, "A and / or B" can mean: only A, only B, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or the like means any combination of these items, including single or multiple combinations. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0198] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the above units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be omitted or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed objects can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0199] The units described above as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0200] In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0201] When the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in part, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes multiple instructions used to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods in the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various other media that can store programs.

[0202] The preferred embodiments of the embodiments of the present application are described above with reference to the accompanying drawings, and are not intended to limit the scope of the embodiments of the present application. Any modifications, equivalent replacements and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of the embodiments of the present application.

Claims

1. A method for recommending items based on object behavior, characterized in that, The method includes the following steps: Collect tracking data of multiple objects. The tracking data includes items accessed by the objects within a preset time period and access information corresponding to the items. The access information includes access type and access time. The items visited by the target object are obtained from the data points, and a first set of items is generated, wherein the target object is one of the plurality of objects; Based on the similarity between the items in the first item set, N candidate recommended items are determined, where N is a preset value greater than 1; Each candidate recommended item is traversed, and the preference weight coefficient of the currently traversed candidate recommended item for the target object is calculated according to the first preset algorithm; Based on the access type of the candidate recommended items to the target object, determine the behavioral weight coefficient of the target object for the candidate recommended items; Based on the access time of the candidate recommended items to the target object, determine the time decay coefficient of the candidate recommended items for the target object; For each candidate recommended item, a recommendation score is determined based on the candidate recommended item's preference weight coefficient, behavior weight coefficient, time decay coefficient, and second similarity to the target object. Based on the recommendation scores of each candidate recommended item, a target recommended item is determined, and the target recommended item is recommended to the target object. The formula for the first preset algorithm is as follows: ； T represents the preference weight coefficient; The number of times the target object accesses the currently traversed candidate recommended items is 'a'. The number of times the target object accesses all items in the first item set; The 'c' represents the number of times the multiple objects access all the candidate recommended items; The d represents the number of times the multiple objects access the currently traversed candidate recommended items.

2. The method according to claim 1, characterized in that, The step of determining N candidate recommended items based on the similarity between the items in the first item set includes: For any two items in the first item set, a first similarity between the two items is determined based on the number of times each item is accessed by the plurality of objects and the number of common access objects of the two items. For each item in the first item set, a second similarity is obtained based on the sum of the first similarities between the item and other items in the first item set; The items in the first item set are sorted from high to low according to the second similarity to obtain an item sequence; The first N items in the item sequence are identified as candidate recommended items, where N is a preset value greater than 1.

3. The method according to claim 1, characterized in that, The step of determining the time decay coefficient of the candidate recommended item for the target object based on the access time of the candidate recommended item for the target object includes: The time interval parameter is determined based on the access time of the candidate recommended items to the target object; The time decay coefficient is determined according to the second preset algorithm and the time interval parameter. The formula for the second preset algorithm is as follows: ； Wherein, F2 represents the degree of interest corresponding to the time interval parameter; F1 is the preset initial level of interest; α represents the time decay coefficient; t is the time interval parameter; e is a natural constant.

4. The method according to claim 2, characterized in that, The formula for calculating the first similarity is as follows: ； Wherein, i represents the i-th item in the first item set, and j represents the j-th item in the first item set; sim[i][j] represents the first similarity between the i-th item and the j-th item; The cnt[i] represents the number of times the i-th item is accessed by the multiple objects; The cnt[j] represents the number of times the j-th item has been accessed by the multiple objects; The mp[i][j] represents the number of common access objects for the i-th item and the j-th item; The This represents the active penalty coefficient.

5. The method according to claim 4, characterized in that, The formula for calculating the activity penalty coefficient is as follows: ； Wherein, |itemlist| represents the number of items included in the first item set.

6. The method according to any one of claims 1-5, characterized in that, The formula for calculating the recommendation score of the candidate recommended item is as follows: ； Wherein, the score represents the recommendation score of the candidate recommended item; T represents the preference weight coefficient; G represents the behavior weighting coefficient; α represents the time decay coefficient; β represents the second similarity of the candidate recommended items.

7. An item recommendation device based on object behavior, characterized in that, The device includes: The tracking data acquisition module is used to collect tracking data of multiple objects. The tracking data includes items accessed by the objects within a preset time period and access information corresponding to the items. The access information includes access type and access time. The first item set generation module is used to obtain items visited by the target object from the data points and generate a first item set, wherein the target object is one of the plurality of objects; The candidate recommended item acquisition module is used to determine N candidate recommended items based on the similarity between the items in the first item set, where N is a preset value greater than 1; The preference weight coefficient acquisition module is used to traverse each of the candidate recommended items and calculate the preference weight coefficient of the currently traversed candidate recommended item for the target object according to the first preset algorithm. The behavior weight coefficient acquisition module is used to determine the behavior weight coefficient of the target object for the candidate recommended item based on the access type of the candidate recommended item for the target object; The time decay coefficient acquisition module is used to determine the time decay coefficient of the candidate recommended item for the target object based on the access time of the candidate recommended item for the target object; The recommendation score acquisition module is used to determine the recommendation score of each candidate recommendation item based on the preference weight coefficient, behavior weight coefficient, time decay coefficient and second similarity of the candidate recommendation item to the target object. The item recommendation module is used to determine the target recommended item based on the recommendation scores of each candidate recommended item, and recommend the target recommended item to the target object; The formula for the first preset algorithm is as follows: ； T represents the preference weight coefficient; The number of times the target object accesses the currently traversed candidate recommended items is 'a'. The number of times the target object accesses all items in the first item set; The 'c' represents the number of times the multiple objects access all the candidate recommended items; The d represents the number of times the multiple objects access the currently traversed candidate recommended items.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the item recommendation method based on object behavior as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements an item recommendation method based on object behavior as described in any one of claims 1 to 6.

Images