Information recommendation method and device, computer device, and storage medium

Abstract

This invention discloses an information recommendation method, apparatus, computer device, and medium, comprising: acquiring registration data of a target object and generating initial tags for the target object based on the registration data; collecting behavioral data of the target object and determining interest words of the target object based on the behavioral data; performing clustering processing based on the initial tags and interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster; acquiring the corresponding tracking points of the clusters and keywords in a preset tracking point set, respectively, as first-level tracking points and second-level tracking points; generating dynamic tracking points based on the first-level tracking points and second-level tracking points, and recommending information to the target object based on the dynamic tracking points. This invention improves the accuracy and efficiency of information recommendation.

Description

Technical Field

[0001] This invention relates to the field of data processing, and more particularly to an information recommendation method, apparatus, computer equipment, and medium. Background Technology

[0002] The internet provides users with abundant information resources. With the rapid development of internet technology, more and more users are obtaining information through the internet. Different groups of people have different target interests. In order to make accurate information recommendations to users, it is necessary to obtain users' interests and preferences.

[0003] Currently, user interests and preferences are mainly obtained through event tracking, which involves counting the frequency of user clicks on various event tracking applications to determine user preferences. However, this method is limited by the number and distribution of the event tracking applications themselves. When the number of event tracking applications is large, the backend needs to monitor a large number of event tracking messages at the same time, which consumes a lot of resources. At the same time, it also results in too much content that users are not interested in, making the information recommendation less targeted. When the number of event tracking applications is small, it is easy to miss user preference information, resulting in less accurate information recommendation. Summary of the Invention

[0004] This invention provides an information recommendation method, apparatus, computer device, and storage medium to improve the accuracy of current information recommendations.

[0005] To address the aforementioned technical problems, embodiments of this application provide an information recommendation method, including:

[0006] Obtain the registration data of the target object, and generate the initialization tag of the target object based on the registration data;

[0007] Collect behavioral data of the target object, and determine the target object's interest words based on the behavioral data;

[0008] Clustering is performed based on the initialization tags and the interest words to obtain at least two clusters, and keywords corresponding to the word vectors contained in each cluster;

[0009] Obtain the corresponding tracking points for the clusters and keywords in the preset tracking point set, respectively, as first-level tracking points and second-level tracking points;

[0010] Based on the primary and secondary tracking points, dynamic tracking points are generated, and information recommendations are made for the target object based on the dynamic tracking points.

[0011] Optionally, determining the target object's interest words based on the behavioral data includes:

[0012] The behavioral data is analyzed and filtered to obtain user keywords;

[0013] The user keywords are trained using word vectors to obtain initial word vectors;

[0014] For each initial word vector, calculate the Euclidean distance from the initial word vector to the word vector corresponding to each preset interest word, and take the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

[0015] Optionally, the clustering process based on the initialization tags and the interest words to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster include:

[0016] The initialization tags are converted into word vectors to obtain the first word vector, and the interest words are converted into word vectors to obtain the second word vector;

[0017] Obtain preset weight information, and based on the preset weight information, perform weighted processing on the first word vector and the second word vector respectively to obtain the updated first word vector and the updated second word vector;

[0018] The K-Means aggregation algorithm is used to cluster the updated first word vector and the updated second word vector to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster.

[0019] Optionally, obtaining the corresponding tracking points for the cluster and the keyword in the preset tracking point set, respectively, as primary tracking points and secondary tracking points, includes:

[0020] Each cluster is designated as a primary tracking point in the preset tracking point set, and each keyword is designated as an initial secondary tracking point in the preset tracking point set.

[0021] For each of the first-level tracking points, a tracking point information heatmap is generated using an iframe based on the distance between keywords in the corresponding cluster of the first-level tracking point.

[0022] Based on the heatmap of the embedded point information and the preset selection method, the initial secondary embedded points are selected to obtain the secondary embedded points.

[0023] Optionally, generating dynamic tracking points based on the primary tracking points and the secondary tracking points includes:

[0024] Construct a mapping relationship between the first-level tracking points and the second-level tracking points, and set the visibility attribute of the first-level tracking points to visible, and set the visibility attribute of the second-level tracking points to invisible;

[0025] When any of the aforementioned first-level tracking points is detected to be triggered, the triggered first-level tracking point is taken as the target first-level tracking point;

[0026] Based on the mapping relationship, the second-level embedding point corresponding to the target first-level embedding point is obtained as the target second-level embedding point;

[0027] Update the visual attribute of the target secondary tracking point to be visible so that the target secondary tracking point is rendered and displayed on the client.

[0028] To address the aforementioned technical problems, embodiments of this application also provide an information recommendation device, comprising:

[0029] The tag acquisition module is used to acquire the registration data of the target object and generate the initial tags of the target object based on the registration data;

[0030] The interest word determination module is used to collect behavioral data of the target object and determine the interest words of the target object based on the behavioral data.

[0031] A clustering processing module is used to perform clustering processing based on the initialization label and the interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster;

[0032] The event tracking layering module is used to obtain the event tracking points corresponding to the clusters and keywords in the preset event tracking set, as first-level event tracking points and second-level event tracking points, respectively.

[0033] The information recommendation module is used to generate dynamic tracking points based on the first-level tracking points and the second-level tracking points, and to recommend information to the target object based on the dynamic tracking points.

[0034] Optionally, the interest word determination module includes:

[0035] The data analysis unit is used to analyze and filter the behavioral data to obtain user keywords;

[0036] The word vector generation unit is used to train the user keywords using word vectors to obtain initial word vectors;

[0037] The interest word determination unit is used to calculate the Euclidean distance from each initial word vector to the word vector corresponding to each preset interest word, and to take the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

[0038] Optionally, the clustering processing module includes:

[0039] The word vector conversion unit is used to convert the initialization tag into a word vector to obtain a first word vector, and to convert the interest word into a word vector to obtain a second word vector;

[0040] The word vector update unit is used to obtain preset weight information, and based on the preset weight information, to perform weighted processing on the first word vector and the second word vector respectively, to obtain the updated first word vector and the updated second word vector.

[0041] A clustering unit is used to cluster the updated first word vector and the updated second word vector using the K-Means aggregation algorithm to obtain at least two clusters, and keywords corresponding to the word vectors contained in each cluster.

[0042] Optionally, the embedding point layering module includes:

[0043] The initial tracking point selection module is used to select the tracking points corresponding to each cluster in the preset tracking point set as first-level tracking points, and to select the tracking points corresponding to each keyword in the preset tracking point set as initial second-level tracking points.

[0044] The heatmap generation module is used to generate a heatmap of the tracking point information for each of the first-level tracking points, based on the distance of keywords in the cluster corresponding to the first-level tracking point, using an iframe.

[0045] The secondary tracking point determination module is used to select the initial secondary tracking points based on the tracking point information heatmap and a preset selection method to obtain the secondary tracking points.

[0046] Optionally, the information recommendation module includes:

[0047] The mapping unit is used to construct the mapping relationship between the first-level tracking point and the second-level tracking point, and to set the visibility attribute of the first-level tracking point to visible and the visibility attribute of the second-level tracking point to invisible.

[0048] The trigger detection unit is used to take the triggered first-level embedding point as the target first-level embedding point when it is detected that any of the first-level embedding points is triggered.

[0049] The target determination unit is used to obtain the second-level embedding point corresponding to the first-level embedding point of the target based on the mapping relationship, and use it as the second-level embedding point of the target;

[0050] The attribute update unit is used to update the visual attributes of the target secondary tracking point to be visible, so that the target secondary tracking point is rendered and displayed on the client.

[0051] To address the aforementioned technical problems, this application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned information recommendation method.

[0052] To address the aforementioned technical problems, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned information recommendation method.

[0053] The information recommendation method, apparatus, computer equipment, and storage medium provided in this invention obtain registration data of a target object and generate initial tags for the target object based on the registration data; collect behavioral data of the target object and determine interest words of the target object based on the behavioral data; perform clustering processing based on the initial tags and interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster; obtain the corresponding tracking points of the clusters and keywords in a preset tracking point set, as first-level tracking points and second-level tracking points; generate dynamic tracking points based on the first-level tracking points and second-level tracking points, and perform information recommendation for the target object based on the dynamic tracking points, thereby realizing rapid personalized and accurate information recommendation for users and improving the accuracy and efficiency of information recommendation. Attached Figure Description

[0054] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;

[0056] Figure 2 This is a flowchart of an embodiment of the information recommendation method of this application;

[0057] Figure 3 This is a schematic diagram of the structure of one embodiment of the information recommendation device according to this application;

[0058] Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0059] 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 this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0060] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see Figure 1 ,like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0063] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc.

[0064] Terminal devices 101, 102, and 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Eperts Group Audio Layer III), MP4 players (Moving Picture Eperts Group Audio Layer IV), laptops, and desktop computers, etc.

[0065] Server 105 can be a server that provides various services, such as a backend server that supports the pages displayed on terminal devices 101, 102, and 103.

[0066] It should be noted that the information recommendation method provided in this application embodiment is executed by the server, and correspondingly, the information recommendation device is set in the server.

[0067] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included. The terminal devices 101, 102, and 103 in this embodiment can specifically correspond to application systems in actual production.

[0068] Please see Figure 2 , Figure 2 This invention illustrates an information recommendation method provided by an embodiment of the present invention, which is applied to... Figure 1 Taking the server-side as an example, the details are as follows:

[0069] S201: Obtain the registration data of the target object and generate the initialization tag of the target object based on the registration data.

[0070] Specifically, when the target (i.e., the client's user) registers, registration data is generated based on the registration information. Based on the user's basic information, registration purpose, personal preference categories, occupation, and other information contained in the registration data, a user profile is created, and initial tags are generated.

[0071] S202: Collect behavioral data of the target object and determine the target object's interest words based on the behavioral data.

[0072] S203: Perform clustering based on the initialization labels and interest words to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster.

[0073] For details, please refer to the description of the following embodiments; to avoid repetition, it will not be repeated here.

[0074] S204: Obtain the corresponding tracking points for clusters and keywords in the preset tracking point set, respectively, as first-level tracking points and second-level tracking points.

[0075] Specifically, the server pre-selects some targeted event points for the target user to form a preset event point set. First-level event points and second-level event points belong to different levels, and a first-level event point includes one or more second-level event points.

[0076] S205: Based on primary and secondary tracking points, generate dynamic tracking points, and based on the dynamic tracking points, recommend information to the target object.

[0077] Specifically, by constructing a mapping relationship between first-level and second-level tracking points, first-level tracking points are set to be visible and second-level tracking points are set to be invisible. When a first-level tracking point is triggered, based on the mapping relationship, the visibility attribute of the second-level tracking point corresponding to the triggered first-level tracking point is set to be visible, so that the second-level tracking point is rendered and displayed on the client.

[0078] In this embodiment, registration data of the target object is obtained, and initial tags for the target object are generated based on the registration data; behavioral data of the target object is collected, and interest words of the target object are determined based on the behavioral data; clustering is performed based on the initial tags and interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster; the corresponding tracking points of the clusters and keywords in the preset tracking point set are obtained respectively, as first-level tracking points and second-level tracking points; dynamic tracking points are generated based on the first-level tracking points and second-level tracking points, and information is recommended to the target object based on the dynamic tracking points, so as to realize rapid personalized and accurate information recommendation for users, and improve the accuracy and efficiency of information recommendation.

[0079] In one specific optional implementation, step S202, determining the target object's interest words based on behavioral data, includes:

[0080] Analyze and filter behavioral data to obtain user keywords;

[0081] Initial word vectors are obtained by training user keywords using word vectors.

[0082] For each initial word vector, calculate the Euclidean distance from the initial word vector to the word vector corresponding to each preset interest word, and take the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

[0083] Specifically, the server retrieves log information recording users' historical behavior from the backend database, and obtains users' behavioral data by extracting the logs.

[0084] User history behavior refers to all user actions after logging into the website platform, which are recorded in the log information of the server's backend database.

[0085] Behavioral data includes, but is not limited to: historical search records, historical click records, and historical browsing records.

[0086] Historical search records include, but are not limited to: user information, search time, and search keywords. User information includes basic user information such as name, gender, and age; search time refers to the specific time when the search operation was detected; and search keywords refer to the keywords entered and used in the search at that time.

[0087] For example, in one specific implementation, the historical search record is: "(Zhang San, male, 23), 2018-01-29 20:46:50, washing machine", where "Zhang San, male, 23" is user information, "2018-01-29 20:46:50" is the search time, and "washing machine" is the search keyword.

[0088] Historical click records include, but are not limited to: user information, click time, and the identification (ID) of the clicked hyperlink. Click time refers to the specific time when the click operation was detected, and the ID of the clicked hyperlink refers to the ID of the object clicked. Furthermore, the click history also records whether the product was clicked in the search results or in the recommended results.

[0089] For example, in one specific implementation, the historical click record is: "(Zhang San, male, 23), 2018-01-29 20:51:50, 65936, S", where "65936" is the ID of the hyperlink on the page providing a detailed introduction to a domestic washing machine in the background database, and "S" is the ID of the clicked hyperlink, which is the ID of the hyperlink obtained by the search method.

[0090] Historical browsing records include, but are not limited to: user information, browsing time, and browsing data. Among these, browsing refers to the browsing records generated when a user browses product information. A single click to view a page will record multiple browsing data entries, which are sampled data.

[0091] Furthermore, by extracting and analyzing key information from the search keywords, the page information corresponding to the clicked hyperlink IDs, and the page information viewed in the behavioral data, the initial keywords of the target object (i.e., the user) can be obtained.

[0092] The page information corresponding to the ID of the clicked hyperlink refers to the page information in the backend database corresponding to the ID of the object being clicked.

[0093] For example, after a user clicks the hyperlink with the ID "NZ_5263", the system queries the backend database to find that the page corresponding to the hyperlink with the ID "NZ_5263" is "https: / / miaosha.xxxxxx.com / #1892018". Then, it extracts the product title, basic product information, etc. of the page to obtain the initial user keywords: "Exquisite and luxurious rhinestone embroidery comparable to the new Korean style knitwear of XXX well-known brand in autumn and winter".

[0094] Furthermore, distracting words are removed from the initial user keywords to obtain the user keywords.

[0095] Specifically, the method for removing distracting words from the user's initial keywords is as follows:

[0096] The initial user keywords are segmented according to a preset dictionary to obtain multiple segments of the initial user keywords;

[0097] Each word segment is classified according to its part of speech. For example, these words are divided into subjective words and objective words according to their part of speech.

[0098] Obtain word segments that meet the preset part-of-speech criteria as user keywords.

[0099] For example, in one specific implementation, the extracted initial user keyword is "exquisite and luxurious rhinestone embroidery comparable to the new Korean style knitwear of XXX well-known brand in autumn and winter". When performing word segmentation, the product title "exquisite and luxurious rhinestone embroidery comparable to the new Korean style knitwear of XXX well-known brand in autumn and winter" can be divided into seven keywords: "exquisite and luxurious", "rhinestone embroidery", "comparable", "XXX well-known brand", "autumn and winter", "new style", "Korean style", and "knitwear". The three words "exquisite and luxurious", "rhinestone embroidery", and "comparable" are subjective words and can be removed. "XXX well-known brand", "autumn and winter", "new style", "Korean style", and "knitwear" are words that can reflect the product characteristics that users need. Therefore, the final five keywords are: "XXX well-known brand", "autumn and winter", "new style", "Korean style", and "knitwear".

[0100] Optionally, when the behavioral information is long, it can be segmented into words and keywords can be extracted. Word segmentation can be achieved using third-party word segmentation tools or word segmentation algorithms.

[0101] Common third-party word segmentation tools include, but are not limited to: Stanford NLP word segmenter, ICTClAS word segmentation system, ansj word segmentation tool, and HanLP Chinese word segmentation tool.

[0102] Among them, word segmentation algorithms include, but are not limited to: Maximum Matching (MM) algorithm, Reverse Direction Maximum Matching Method (RMM) algorithm, Bi-direction Matching method (BM) algorithm, Hidden Markov Model (HMM) and N-gram model, etc.

[0103] Furthermore, word vectors corresponding to behavioral information are generated. Specifically, each keyword corresponding to the generated behavioral information is mapped to a vector according to a preset corpus. These vectors are then linked together to form a word vector space, where each vector is equivalent to a point in this space. Each vector is treated as a spatial word vector.

[0104] In one specific optional implementation, in step S203, clustering is performed based on the initialization tags and interest words to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster include:

[0105] The initial tags are converted into word vectors to obtain the first word vector, and the interest words are converted into word vectors to obtain the second word vector;

[0106] Obtain the preset weight information, and based on the preset weight information, perform weighted processing on the first word vector and the second word vector respectively to obtain the updated first word vector and the updated second word vector;

[0107] The K-Means aggregation algorithm is used to cluster the updated first word vector and the updated second word vector to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster.

[0108] The K-means algorithm is a distance-based clustering algorithm that uses distance as a similarity metric, meaning that the closer two objects are, the greater their similarity. This algorithm considers clusters to be composed of objects that are close together, thus aiming to obtain compact and independent clusters as its ultimate goal.

[0109] The preset weight information can be set according to actual needs. The preset weight information corresponding to the initialization tag and the preset weight information corresponding to the interest word can be the same or different. There is no specific limitation here.

[0110] In this embodiment, the K-Means aggregation algorithm is used to perform cluster analysis on the updated first word vector and the updated second word vector, which is described in detail below:

[0111] Initialize m cluster centers, where m is the preset number of cluster centers;

[0112] For each updated first word vector and each updated second word vector, calculate the first distance between the spatial word vector and each current cluster center, and put the spatial word vector into the cluster where the cluster center with the smallest first distance is located, to obtain m temporary clusters;

[0113] For each temporary cluster, calculate the mean of the temporary cluster and the second distance between each spatial word vector in the temporary cluster and the mean. Select the spatial word vector corresponding to the smallest second distance as the new cluster center of the temporary cluster, and obtain the updated m temporary clusters.

[0114] Calculate the standard deviation of each updated temporary cluster using the following formula:

[0115]

[0116] Where σ is the standard deviation, A i Let A be the i-th spatial word vector in the updated temporary cluster, n be the number of spatial word vectors in the updated temporary cluster, and μ be the spatial word vector A. i The mean of the updated temporary cluster in which it is located, i∈[1,n], and i and n are positive integers;

[0117] If at least one of the standard deviations of the m updated temporary clusters is greater than or equal to the preset standard deviation threshold, then return to the step of performing the following for each spatial word vector in the effective evaluation: calculate the first distance between the spatial word vector and each current cluster center, and put the spatial word vector into the cluster where the cluster center with the smallest first distance is located, thus obtaining m temporary clusters;

[0118] If the standard deviation of all m updated temporary clusters is less than the standard deviation threshold, then the cluster centers of the m updated temporary clusters are taken as the cluster analysis result. The cluster analysis result includes at least m clusters, and the keywords corresponding to the word vectors contained in each cluster.

[0119] In one specific optional implementation, step S204, obtaining the corresponding tracking points for clusters and keywords in a preset tracking point set, as primary tracking points and secondary tracking points respectively, includes:

[0120] Each cluster is designated as a primary tracking point in the preset tracking point set, and each keyword is designated as an initial secondary tracking point in the preset tracking point set.

[0121] For each primary tracking point, a heatmap of tracking point information is generated using an iframe based on the distance between keywords in the corresponding cluster.

[0122] Based on the heat map of the embedded point information and the preset selection method, the initial secondary embedded points are selected to obtain the secondary embedded points.

[0123] The IFRAME is an HTML tag used to create a document within a document, or a floating frame. The iframe element creates an inline frame that contains another document.

[0124] Heatmaps are a commonly used display method in data visualization projects. They visually reflect data information such as hotspot distribution and regional clustering through color variations. This embodiment uses a heatmap of event tracking information to quickly identify popular or highly-watched secondary event tracking points. When there are many secondary event tracking points, popular or highly-watched points are prioritized for recommendation, which helps improve the accuracy of recommendations.

[0125] In one specific optional implementation, step S205, generating dynamic tracking points based on primary tracking points and secondary tracking points, includes:

[0126] Establish a mapping relationship between first-level and second-level tracking points, and set the visibility attribute of first-level tracking points to visible, and set the visibility attribute of second-level tracking points to invisible;

[0127] When any level-one tracking point is detected to be triggered, the triggered level-one tracking point will be used as the target level-one tracking point.

[0128] Based on the mapping relationship, obtain the second-level tracking points corresponding to the target first-level tracking points, and use them as the target second-level tracking points;

[0129] Update the visual attribute of the target secondary tracking point to "visible" so that the target secondary tracking point is rendered and displayed on the client.

[0130] When a level-one tracking point is triggered, tracking point trigger information is generated. Specifically, the tracking point trigger information can be written to a background log file and retrieved from the background log file periodically, or a trigger message containing tracking point trigger information can be generated and fed back to the server when the tracking point is triggered.

[0131] The visibility attribute refers to whether the data point can be displayed on the client side, including visible and invisible. When the visibility attribute is visible, the data point can be rendered and displayed by the client side; when the visibility attribute is invisible, the data point cannot be rendered and displayed by the client side.

[0132] In this embodiment, dynamic tracking points are generated through primary and secondary tracking points to improve the accuracy of locating user preferences, which is beneficial to improving the precision of information recommendation.

[0133] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0134] Figure 3 A schematic block diagram of an information recommendation device corresponding to the information recommendation method in the above embodiments is shown. For example... Figure 3 As shown, the information recommendation device includes a tag acquisition module 31, an interest word determination module 32, a clustering processing module 33, a data tracking and layering module 34, and an information recommendation module 35. Detailed descriptions of each functional module are as follows:

[0135] The tag acquisition module 31 is used to acquire the registration data of the target object and generate the initialization tag of the target object based on the registration data;

[0136] The interest word determination module 32 is used to collect behavioral data of the target object and determine the interest words of the target object based on the behavioral data;

[0137] Clustering module 33 is used to perform clustering based on initialization labels and interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster;

[0138] The event tracking layering module 34 is used to obtain the corresponding event tracking points of clusters and keywords in the preset event tracking set, as first-level event tracking points and second-level event tracking points, respectively.

[0139] The information recommendation module 35 is used to generate dynamic tracking points based on primary and secondary tracking points, and to recommend information to target objects based on these dynamic tracking points.

[0140] Optionally, the interest word determination module 32 includes:

[0141] The data analysis unit is used to analyze and filter behavioral data to obtain user keywords;

[0142] The word vector generation unit is used to train user keywords using word vectors to obtain initial word vectors;

[0143] The interest word determination unit is used to calculate the Euclidean distance from each initial word vector to the word vector corresponding to each preset interest word, and to select the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

[0144] Optionally, the clustering processing module 33 includes:

[0145] The word vector conversion unit is used to convert the initial tags into word vectors to obtain the first word vector, and to convert the interest words into word vectors to obtain the second word vector;

[0146] The word vector update unit is used to obtain preset weight information and, based on the preset weight information, perform weighted processing on the first word vector and the second word vector respectively to obtain the updated first word vector and the updated second word vector.

[0147] Clustering units are used to cluster the updated first word vector and the updated second word vector using the K-Means aggregation algorithm to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster.

[0148] Optionally, the data embedding layer module 34 includes:

[0149] The initial tracking point selection module is used to select the tracking points corresponding to each cluster in the preset tracking point set as first-level tracking points, and to select the tracking points corresponding to each keyword in the preset tracking point set as initial second-level tracking points.

[0150] The heatmap generation module is used to generate a heatmap of the tracking point information for each primary tracking point, based on the distance between keywords in the corresponding cluster of the primary tracking point and the tracking point, using an iframe.

[0151] The secondary tracking point determination module is used to select initial secondary tracking points based on the tracking point information heatmap and preset selection methods to obtain secondary tracking points.

[0152] Optionally, the information recommendation module 35 includes:

[0153] The mapping unit is used to construct the mapping relationship between first-level and second-level tracking points, and set the visibility attribute of the first-level tracking points to visible, and the visibility attribute of the second-level tracking points to invisible.

[0154] The trigger detection unit is used to take the triggered first-level tracking point as the target first-level tracking point when any first-level tracking point is detected to be triggered.

[0155] The target determination unit is used to obtain the second-level tracking points corresponding to the first-level tracking points of the target based on the mapping relationship, and use them as the second-level tracking points of the target;

[0156] The attribute update unit is used to update the visible attributes of the target secondary tracking point to be visible, so that the target secondary tracking point can be rendered and displayed on the client.

[0157] Specific limitations regarding the information recommendation device can be found in the limitations of the information recommendation method described above, and will not be repeated here. Each module in the aforementioned information recommendation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0158] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0159] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components connected to the memory 41, processor 42, and network interface 43 is shown in the figure; however, it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0160] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0161] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or D-interface display memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as program code for controlling electronic documents. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0162] In some embodiments, the processor 42 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run program code stored in the memory 41 or process data, such as program code controlling electronic documents.

[0163] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0164] This application also provides another embodiment, namely, a computer-readable storage medium storing an interface display program that can be executed by at least one processor to cause the at least one processor to perform the steps of the information recommendation method described above.

[0165] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0166] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. An information recommendation method, characterized in that, The information recommendation method includes: Obtain the registration data of the target object, and generate the initialization tag of the target object based on the registration data; Collect behavioral data of the target object, and determine the target object's interest words based on the behavioral data; Clustering is performed based on the initialization tags and the interest words to obtain at least two clusters, and keywords corresponding to the word vectors contained in each cluster; Obtain the corresponding tracking points for the clusters and keywords in the preset tracking point set, respectively, as first-level tracking points and second-level tracking points; Based on the primary and secondary tracking points, dynamic tracking points are generated, and a mapping relationship between the primary and secondary tracking points is constructed. The visibility attribute of the primary tracking points is set to visible, and the visibility attribute of the secondary tracking points is set to invisible. When any primary tracking point is detected to be triggered, the triggered primary tracking point is taken as the target primary tracking point. Based on the mapping relationship, the secondary tracking point corresponding to the target primary tracking point is obtained and taken as the target secondary tracking point. The visibility attribute of the target secondary tracking point is updated to visible so that the target secondary tracking point is rendered and displayed on the client. Based on the dynamic tracking points, information recommendation is performed on the target object.

2. The information recommendation method as described in claim 1, characterized in that, The step of determining the target object's interest words based on the behavioral data includes: The behavioral data is analyzed and filtered to obtain user keywords; The user keywords are trained using word vectors to obtain initial word vectors; For each initial word vector, calculate the Euclidean distance from the initial word vector to the word vector corresponding to each preset interest word, and take the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

3. The information recommendation method as described in claim 1, characterized in that, The clustering process based on the initialization tags and the interest words yields at least two clusters, and the keywords corresponding to the word vectors contained in each cluster include: The initialization tags are converted into word vectors to obtain the first word vector, and the interest words are converted into word vectors to obtain the second word vector; Obtain preset weight information, and based on the preset weight information, perform weighted processing on the first word vector and the second word vector respectively to obtain the updated first word vector and the updated second word vector; The K-Means aggregation algorithm is used to cluster the updated first word vector and the updated second word vector to obtain at least two clusters, and the keywords corresponding to the word vectors contained in each cluster.

4. The information recommendation method as described in claim 1, characterized in that, The step of obtaining the corresponding tracking points of the cluster and the keyword in the preset tracking point set, as primary tracking points and secondary tracking points respectively, includes: Each cluster is designated as a primary tracking point in the preset tracking point set, and each keyword is designated as an initial secondary tracking point in the preset tracking point set. For each of the first-level tracking points, a tracking point information heatmap is generated using an iframe based on the distance between keywords in the corresponding cluster of the first-level tracking point. Based on the heatmap of the embedded point information and the preset selection method, the initial secondary embedded points are selected to obtain the secondary embedded points.

5. An information recommendation device, characterized in that, The information recommendation device includes: The tag acquisition module is used to acquire the registration data of the target object and generate the initial tags of the target object based on the registration data; The interest word determination module is used to collect behavioral data of the target object and determine the interest words of the target object based on the behavioral data. A clustering processing module is used to perform clustering processing based on the initialization label and the interest words to obtain at least two clusters, and keywords corresponding to word vectors contained in each cluster; The event tracking layering module is used to obtain the event tracking points corresponding to the clusters and keywords in the preset event tracking set, as first-level event tracking points and second-level event tracking points, respectively. The information recommendation module is used to generate dynamic tracking points based on the primary tracking points and the secondary tracking points, construct a mapping relationship between the primary tracking points and the secondary tracking points, set the visibility attribute of the primary tracking points to visible, and set the visibility attribute of the secondary tracking points to invisible; when any of the primary tracking points is detected to be triggered, the triggered primary tracking point is taken as the target primary tracking point; based on the mapping relationship, the secondary tracking point corresponding to the target primary tracking point is obtained as the target secondary tracking point; the visibility attribute of the target secondary tracking point is updated to visible so that the target secondary tracking point is rendered and displayed on the client, and information recommendation is performed on the target object based on the dynamic tracking points.

6. The information recommendation device as described in claim 5, characterized in that, The interest word determination module includes: The data analysis unit is used to analyze and filter the behavioral data to obtain user keywords; The word vector generation unit is used to train the user keywords using word vectors to obtain initial word vectors; The interest word determination unit is used to calculate the Euclidean distance from each initial word vector to the word vector corresponding to each preset interest word, and to take the preset interest word corresponding to the smallest Euclidean distance as the interest word of the target object.

7. The information recommendation device as described in claim 5, characterized in that, The The clustering processing module includes: The word vector conversion unit is used to convert the initialization tag into a word vector to obtain a first word vector, and to convert the interest word into a word vector to obtain a second word vector; The word vector update unit is used to obtain preset weight information and, based on the preset weight information, perform weighted processing on the first word vector and the second word vector respectively to obtain the updated first word vector and the updated second word vector. A clustering unit is used to cluster the updated first word vector and the updated second word vector using the K-Means aggregation algorithm to obtain at least two clusters, and keywords corresponding to the word vectors contained in each cluster.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the information recommendation method as described in any one of claims 1 to 4.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the information recommendation method as described in any one of claims 1 to 4.

Images