A method, apparatus, computer device, and storage medium for information push.
By pushing targeted multimedia content and search terms into search engines, the problem of users' lack of understanding of search engines is solved, and the utilization rate of search engine resources and users' search efficiency are improved.
Patent Information
- Application Number
- CN202211213721.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-30
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2042-09-30
AI Technical Summary
Users cannot know in advance what kind of content a search engine can search, resulting in wasted search engine resources and low efficiency.
By acquiring multimedia content and search terms from the resource library, we can identify target multimedia content and target search terms that can help users build search awareness, and then push them through different channels to help users understand the relationship between search behavior and search results.
It improves the utilization rate of search engine-related resources, helping users to better use search engines for information retrieval.
Smart Images

Figure CN115577126B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of computer technology, and more specifically, to a method, apparatus, computer device, and storage medium for information push. Background Technology
[0002] Search engines provide users with information retrieval services. Users can enter search terms on a search page and use the search engine to search for information matching those terms from a resource database. Currently, search engines are widely used in various application scenarios, especially in multimedia platforms that provide multimedia content. Users can use search engines to search for specific types and content of multimedia content from massive amounts of data.
[0003] To fully utilize the functions of a search engine and find multimedia content that matches your search intent, users need to have a certain understanding of the search engine's capabilities. This means understanding what aspects of content a search engine can search for and what search capabilities it can provide. This understanding will directly affect search results and the efficiency of the search engine's resource utilization. Summary of the Invention
[0004] This disclosure provides at least one method, apparatus, computer device, and storage medium for information push.
[0005] In a first aspect, embodiments of this disclosure provide a method for information push, including:
[0006] Retrieve multimedia content and search terms from the resource library;
[0007] From the multimedia content, identify target multimedia content that can establish search awareness, and from the search terms, identify target search terms that can establish search awareness;
[0008] The target multimedia content and the target search terms will be pushed through different channels.
[0009] In one optional implementation, the target multimedia content and the target search term are pushed through different channels, including:
[0010] Based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category is determined; the set of distribution objects includes multiple target distribution objects.
[0011] For the set of distribution objects, a distribution weight is determined for each preset vertical category; wherein the distribution weight of the preset vertical category that matches the set of distribution objects is greater than the distribution weight of other preset vertical categories; the distribution weight is used to indicate the proportion of the distribution quantity of the corresponding preset vertical category in each preset vertical category;
[0012] According to the distribution weights corresponding to each preset vertical category, the target multimedia content and the target search terms are distributed to each target distribution object in the distribution object set through different channels.
[0013] In one optional implementation, based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category is determined, including:
[0014] For each of the preset vertical categories, the same amount of initial distribution information is determined, wherein the initial distribution information includes the target multimedia content to be initially distributed, and / or the target search terms to be initially distributed;
[0015] Distribute the initial distribution information to each distribution object;
[0016] Based on the browsing data corresponding to each preset vertical category within a preset time period after distribution, a set of distribution objects matching each preset vertical category is determined.
[0017] In another optional implementation, the target multimedia content and the target search term are pushed through different channels, including:
[0018] For each preset vertical category, based on the attribute dimensions matched by the preset vertical category, the importance scores of each target multimedia content and the target search term under the preset vertical category are determined;
[0019] Based on the distribution status of the preset vertical category on each distribution channel, and the importance scores of each target multimedia content and target search term under the preset vertical category, the target multimedia content and target search term to be distributed are selected for each preset vertical category.
[0020] The target multimedia content and target search terms selected for each of the preset vertical categories will be pushed through different channels.
[0021] In one optional implementation, determining target multimedia content capable of establishing search awareness from the multimedia content, and determining target search terms capable of establishing search awareness from the search terms, includes:
[0022] Based on multiple search cognitive dimensions, initial multimedia content matching each search cognitive dimension is determined from the multimedia content, and initial search terms matching each search cognitive dimension are determined from the search terms; the search cognitive dimension refers to the category dimension corresponding to the multimedia content that can establish search cognition.
[0023] For each of the search cognitive dimensions, feature clustering is performed on the initial multimedia content and initial search terms to determine at least one first feature condition and at least one second feature condition corresponding to each of the search cognitive dimensions; the first feature condition is matched with the multimedia content, and the second feature condition is matched with the search terms.
[0024] The target multimedia content that meets the first feature condition is determined from the multimedia content, and the target search term that meets the second feature condition is determined from the search term.
[0025] In one optional implementation, determining target multimedia content capable of establishing search awareness from the multimedia content, and determining target search terms capable of establishing search awareness from the search terms, includes:
[0026] For each multimedia content, first search feature data before browsing the multimedia content, second search feature data after browsing the multimedia content, and first target feature data of not browsing the multimedia content are obtained; and for each search term, third search feature data before initiating a search based on the search term, fourth search feature data after initiating a search based on the search term, and second target feature data of not initiating a search using the search term are obtained.
[0027] The first search feature data, the second search feature data, and the first target feature data corresponding to each of the multimedia contents are input into the trained first search cognitive model to obtain the first search activity improvement value corresponding to each of the multimedia contents. The third search feature data, the fourth search feature data, and the second target feature data corresponding to each of the search terms are input into the trained second search cognitive model to obtain the second search activity improvement value corresponding to each of the search terms.
[0028] Based on the first search activity increase value and the second search activity increase value, the target multimedia content and the target search term are determined from each of the multimedia content and the search term.
[0029] In one optional implementation, the target multimedia content and the target search term are pushed through different channels, including:
[0030] The target multimedia content is displayed on the information feed recommendation page, and the target search term is displayed on the target information page.
[0031] In one optional implementation, displaying the target search term on the target information page includes:
[0032] The target search term is displayed on at least one of the following pages: the search page where the search was initiated, the search results page, and the multimedia content details page.
[0033] Secondly, embodiments of this disclosure also provide an information push device, comprising:
[0034] The acquisition module is used to retrieve multimedia content and search terms from the resource library;
[0035] The determining module is used to determine, from the multimedia content, target multimedia content that can establish search cognition, and from the search terms, target search terms that can establish search cognition;
[0036] The push module is used to push the target multimedia content and the target search terms through different channels respectively.
[0037] Thirdly, embodiments of this disclosure also provide a computer device, including: a processor and a memory, the memory storing machine-readable instructions executable by the processor, the processor being configured to execute the machine-readable instructions stored in the memory, and when the machine-readable instructions are executed by the processor, the processor performs the steps of the information push method described in any of the preceding aspects.
[0038] Fourthly, embodiments of this disclosure also include a computer-readable storage medium storing a computer program, wherein when the computer program is run by a computer device, the computer device performs the steps of the information push method described in any of the preceding aspects.
[0039] The information push method, apparatus, computer equipment, and storage medium provided in this disclosure can acquire multimedia content and search terms from a resource library, determine target multimedia content that can establish search awareness from the multimedia content, and determine target search terms that can establish search awareness from the search terms. Then, the target multimedia content and target search terms can be pushed through different channels, allowing users to browse the target multimedia content and target search terms. Since the target multimedia content and target search terms have the characteristic of establishing search awareness, they can help users understand the search engine and the relationship between search behavior and search results. Thus, when users have search needs, the search engine can better perform its functions, helping users to better use the search engine for information retrieval while improving the utilization rate of search engine-related resources.
[0040] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0041] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments will be briefly described below. These drawings are incorporated in and constitute a part of this specification. They illustrate embodiments conforming to this disclosure and, together with the specification, serve to explain the technical solutions of this disclosure. It should be understood that the following drawings only show some embodiments of this disclosure and should not be considered as limiting the scope. Those skilled in the art can obtain other related drawings based on these drawings without creative effort.
[0042] Figure 1 A flowchart of an information push method provided by an embodiment of this disclosure is shown;
[0043] Figure 2 This illustration shows a flowchart of a method for pushing target multimedia content and target search terms through different channels, as provided in an embodiment of this disclosure.
[0044] Figure 3 The flowchart illustrates a method for determining a set of distribution objects matching each preset vertical category based on browsing data corresponding to each preset vertical category, as provided in an embodiment of this disclosure.
[0045] Figure 4 A flowchart is shown illustrating another method for pushing target multimedia content and target search terms through different channels, as provided in this disclosure embodiment.
[0046] Figure 5 A schematic diagram of an information push device provided in an embodiment of this disclosure is shown;
[0047] Figure 6 A schematic diagram of a computer device provided in an embodiment of this disclosure is shown. Detailed Implementation
[0048] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. The components of the embodiments of this disclosure described and shown herein can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this disclosure is not intended to limit the scope of the claimed disclosure, but merely represents selected embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of protection of this disclosure.
[0049] Research has shown that search engines are an important tool for users to obtain the information they need. However, users usually do not know in advance what kind of content a search engine can search for, and they need to try searching multiple times to develop an understanding of search engines. This causes some users to miss many opportunities to use search engines to find the information they want, and also leads to a waste of search engine resources.
[0050] The shortcomings of the above solutions are the result of the inventor's practical experience and careful research. Therefore, the discovery process of the above problems and the solutions proposed in this disclosure below should be considered as the inventor's contribution to this disclosure.
[0051] Based on this, this disclosure provides an information push method that can acquire multimedia content and search terms from a resource library, determine target multimedia content that can establish search awareness from the multimedia content, and determine target search terms that can establish search awareness from the search terms; then, the target multimedia content and target search terms can be pushed through different channels, allowing users to browse the target multimedia content and target search terms. Since the target multimedia content and target search terms have the characteristic of establishing search awareness, they can help users understand the search engine and the relationship between search behavior and search results, thereby better leveraging the function of the search engine when users have search needs. While helping users better use the search engine for information retrieval, it also improves the utilization rate of search engine-related resources.
[0052] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.
[0053] To facilitate understanding of this embodiment, a method for pushing information disclosed in this disclosure will first be described in detail. The execution subject of the information pushing method provided in this disclosure is generally a computer device with certain computing capabilities, such as a terminal device, a server, or other processing devices. In some possible implementations, the information pushing method can be implemented by a processor calling computer-readable instructions stored in memory.
[0054] The method for pushing information provided in the embodiments of this disclosure will be described below.
[0055] See Figure 1 The diagram shows a flowchart of an information push method provided in an embodiment of this disclosure. The method includes steps S101 to S103, wherein:
[0056] S101: Retrieve multimedia content and search terms from the resource library.
[0057] S102: From multimedia content, identify target multimedia content that can establish search awareness, and from search terms, identify target search terms that can establish search awareness.
[0058] S103: Push target multimedia content and target search terms through different channels.
[0059] This embodiment of the disclosure, after acquiring multimedia content and search terms from the resource library, can determine target multimedia content that can establish search awareness from the multimedia content, and target search terms that can establish search awareness from the search terms. Then, the target multimedia content and target search terms can be pushed through different channels, allowing users to browse the target multimedia content and target search terms. Since the target multimedia content and target search terms have the characteristic of establishing search awareness, they can help users understand the search engine and the relationship between search behavior and search results. Thus, when users have search needs, the search engine can better perform its functions, helping users to better use the search engine for information retrieval while improving the utilization rate of search engine-related resources.
[0060] The above S101 to S103 are explained in detail below.
[0061] Regarding the above S101:
[0062] The resource library can include various multimedia content and search terms. Specifically, multimedia content can include various types of short videos, movies, TV series, novels, news, etc., such as short game videos, short videos on cooking methods, and romantic novels. Search terms can include movie titles, novel titles or authors, media accounts, etc. The resource library can be the database of the multimedia platform, storing various searchable resources. Multimedia content and search terms can be stored in the same resource library or in separate resource libraries. The multimedia platform can publish a variety of multimedia content and provide a search engine. Users can use the search engine to retrieve multimedia content. Recommended search terms can be displayed on any page where information is displayed on the platform. Users can trigger these recommended search terms or actively enter search terms to conduct searches.
[0063] Regarding S102 above:
[0064] After acquiring multimedia content and search terms, target multimedia content and target search terms that can build search awareness can be identified from the multimedia content and search terms, respectively. For example, a trained model can be used to classify the multimedia content and search terms separately, or the multimedia content and search terms can be analyzed from multiple dimensions to determine the target multimedia content and target search terms that can build search awareness. Building search awareness means that after browsing the target multimedia content or target search terms, users can improve their understanding of how to use the search engine. This is manifested in ways such as increased search frequency, improved search accuracy, and improved search efficiency after browsing the target multimedia content or target search terms.
[0065] In one optional implementation, based on the search cognition dimension, target multimedia content capable of establishing search cognition can be determined from multimedia content, and target search terms capable of establishing search cognition can be determined from search terms, including:
[0066] Based on multiple search cognitive dimensions, the initial multimedia content matching each search cognitive dimension is determined from the multimedia content, and the initial search terms matching each search cognitive dimension are determined from the search terms; the search cognitive dimension refers to the category dimension corresponding to the multimedia content that can establish search cognition.
[0067] For each search cognitive dimension, feature clustering is performed on the initial multimedia content and initial search terms to determine at least one first feature condition and at least one second feature condition for each search cognitive dimension; the first feature condition is matched with the multimedia content, and the second feature condition is matched with the search terms.
[0068] The first feature condition is determined from the multimedia content, and the second feature condition is determined from the search terms.
[0069] Here, the search cognition dimension can refer to the category dimension corresponding to the multimedia content or search terms that can establish search cognition. For example, the search cognition dimension can include the short-term search motivation stimulation dimension, the stimulation of high-value search types dimension, and the generalization ability dimension.
[0070] Among them, the short-term search motivation dimension can characterize the impact of multimedia content or search terms on users' short-term search motivation. Short-term search motivation can be manifested as users' willingness to search for more content in the short term after browsing relevant multimedia content, or users' willingness to use more search terms in the short term after using relevant search terms.
[0071] The dimension of stimulating high-value search types can characterize the post-view search types brought about by multimedia content or search terms. By judging whether the post-view search types under this dimension are high-value types, we can determine whether the multimedia content or search terms meet the standards for establishing search cognition. Here, post-view search type refers to the request type corresponding to the search request initiated after browsing relevant multimedia content or using relevant search terms. High-value type is the request type with high search cognition value. That is, before searching, the user does not know what the result that meets their needs should be. The search result is information that the user does not know. The search behavior is effective, which reflects the user's recognition of the comprehensive search capabilities of the search engine. This kind of multimedia content that allows users to learn more information through searching generally helps to form search cognition.
[0072] The generalization ability dimension can characterize a user's long-term search intention after browsing these multimedia content or search terms. Under the generalization ability dimension, we can analyze whether the search behavior after viewing the multimedia content belongs to generalized search behavior (that is, there is content in the same field but unrelated content in the multimedia content searched after viewing). If it belongs to generalized search behavior, it means that the user can expand on the multimedia content or search terms and has divergent thinking in search.
[0073] During implementation, the initial multimedia content matching each search cognitive dimension can be determined from the multimedia content based on multiple search cognitive dimensions, and the initial search terms matching each search cognitive dimension can be determined from the search terms.
[0074] For example, regarding multimedia content, for the short-term search motivation dimension, the ratio between the number of times each user initiated searches for content related to that multimedia content within a preset time period after watching it, and the number of times each user watched the multimedia content, can be calculated. The average of the ratios for each multimedia content in the resource library can also be calculated. Then, multimedia content with a ratio greater than the average can be selected as the initial multimedia content. Similarly, for search terms, the ratio between the number of times each user initiated searches for content related to that search term within a preset time period after using it, and the number of times each user used that search term, can also be calculated. The average of the ratios for each search term in the resource library can also be calculated. Then, search terms with a ratio greater than the average can be selected as the initial search terms.
[0075] To stimulate high-value search types, we can select multimedia content / search terms matching the target search type from the resource library based on the search type corresponding to each multimedia content / search term. For example, initial multimedia content could include information on trending topics or events, recipes, or encyclopedic knowledge; initial search terms could include recipe collections or mobile app shortcuts. Here, the search type can be understood as the type of request initiated after browsing relevant multimedia content or using relevant search terms.
[0076] Regarding the generalization capability dimension, for each multimedia content / search term, based on the domain category to which the multimedia content / search term and its corresponding search content belong, and the relevance between the multimedia content / search term and its corresponding search content, multimedia content that matches the domain category of the multimedia content and has a relevance below a set threshold can be selected from the resource library as initial multimedia content. For example, if the multimedia content is "how to make tofu pudding sauce", the initial multimedia content could be "how to make sweet and sour pork ribs"; if the search term is "online games", the initial search term could be "single-player games". Here, search content can be understood as the content searched after browsing relevant multimedia content or using relevant search terms.
[0077] In practical implementation, after obtaining the initial multimedia content matching each search cognitive dimension and the initial search terms matching each search cognitive dimension, the characteristics of the multimedia content corresponding to each search cognitive dimension can be summarized based on the initial multimedia content and initial search terms, i.e., at least one first feature condition, and the characteristics of the search terms corresponding to each search cognitive dimension, i.e., at least one second feature condition. Therefore, more target multimedia content that can bring search cognitive can be further filtered based on the first feature condition, and more target search terms that can bring search cognitive can be further filtered based on the second feature condition. The first and second feature conditions can respectively include multimedia content classification, browsing characteristics, and the conversion rate range of searching after viewing, etc.
[0078] In another alternative implementation, based on the model, target multimedia content capable of establishing search awareness can be determined from multimedia content, and target search terms capable of establishing search awareness can be determined from search terms, including:
[0079] For each multimedia content, first search feature data before browsing the multimedia content, second search feature data after browsing the multimedia content, and first target feature data of not browsing the multimedia content are obtained; and for each search term, third search feature data before initiating a search based on the search term, fourth search feature data after initiating a search based on the search term, and second target feature data of not initiating a search using the search term are obtained.
[0080] Input the first search feature data, second search feature data, and first target feature data corresponding to each multimedia content into the trained first search cognitive model to obtain the first search activity improvement value corresponding to each multimedia content; and input the third search feature data, fourth search feature data, and second target feature data corresponding to each search term into the trained second search cognitive model to obtain the second search activity improvement value corresponding to each search term.
[0081] Based on the first search activity increase value and the second search activity increase value, the target multimedia content and target search terms are determined from various multimedia content and search terms.
[0082] Here, for each multimedia content, we can obtain first search feature data within a first preset time period before viewing the multimedia content, second search feature data within a second preset time period after viewing the multimedia content, and first target feature data for multimedia content that has not been viewed within the most recent third preset time period. The first and second preset time periods have the same duration; for example, the first preset time period can be 7 days. The duration of the third preset time period can be the same as or different from the durations of the first and second preset time periods. The first and second search feature data may include at least one of the following: search count, domain type corresponding to the search information, search result quality information, language structure features of the search information, and search channel information. The first target feature data may include at least one of the following: search feature data, multimedia content browsing data, comment data, and usage data of the platform distributing the multimedia content.
[0083] Similar to the steps for determining the initial multimedia content, for each search term, third search feature data can be obtained within a fourth preset time period before the search term was initiated, fourth search feature data within a fifth preset time period after the search term was initiated, and second target feature data for searches not initiated using the search term within the most recent sixth preset time period; wherein the fourth, fifth, and sixth preset time periods correspond to the same duration, for example, the fourth preset time period can be 12 hours; the third and fourth search feature data can include at least one of the following: search count, domain type corresponding to the search information, search result quality information, language structure features of the search information, and search channel information; the second target feature data can include at least one of the following: search feature data, search term usage data, and usage data for the platform that distributed the search term.
[0084] After obtaining the above features, the first search feature data, second search feature data, and first target feature data corresponding to each multimedia content can be input into the trained first search cognitive model to obtain the first search activity improvement value corresponding to each multimedia content; and the third search feature data, fourth search feature data, and second target feature data corresponding to each search term can be input into the trained second search cognitive model to obtain the second search activity improvement value corresponding to each search term. The search activity improvement value can be the average search activity improvement value, which can be used to represent the increase in the ratio of the number of first users who generated the first target search behavior within a second preset time period after browsing the multimedia content to the number of second users who generated the first target search behavior within a first preset time period before browsing the multimedia content; and it can also be used to represent the increase in the ratio of the number of third users who generated the second target search behavior within a fifth preset time period after initiating a search using the search term to the number of fourth users who generated the second target search behavior within a fourth preset time period after initiating a search using the search term.
[0085] The first and second search cognitive models can be pre-trained. Specifically, for the first search cognitive model, we can first obtain the first search feature data sample before browsing the multimedia content sample, the second search feature data sample after browsing the multimedia content sample, and the first target feature data sample that has not been browsed. Then, we use the first search feature data sample, the second search feature data sample, and the first target feature data sample as the input results of the first search cognitive model, and use the increase in average search activity per user after browsing the multimedia content sample compared to before browsing the multimedia content sample as the output result of the first search cognitive model, and train the first search cognitive model.
[0086] For the second search cognitive model, we can first obtain the third search feature data sample before the search is initiated based on the search term sample, the fourth search feature data sample after the search is initiated based on the search term sample, and the second target feature data sample without using the search term sample. Then, we use the third search feature data sample, the fourth search feature data sample, and the second target feature data sample as the input results of the second search cognitive model, and use the increase in average search activity per user after the search is initiated based on the search term sample compared to before the search is initiated based on the search term sample as the output result of the second search cognitive model, and train the second search cognitive model.
[0087] Then, based on the first search activity improvement value corresponding to each multimedia content, initial multimedia content with a first search activity improvement value greater than a first set threshold can be selected from the resource library; and based on the first target feature data corresponding to each multimedia content and the first weight value corresponding to each first target feature data, a first recommendation score corresponding to each initial multimedia content can be determined; furthermore, a first preset number of target multimedia content can be selected from each initial multimedia content in descending order of the first recommendation score; and based on the second search activity improvement value corresponding to each search term, initial search terms with a second search activity improvement value greater than a second set threshold can be selected from the resource library; and based on the second target feature data corresponding to each search term and the second weight value corresponding to each second target feature data, a second recommendation score corresponding to each initial search term can be determined; furthermore, a second preset number of target search terms can be selected from each initial search term in descending order of the second recommendation score. The first set threshold, the second set threshold, the first preset number, and the second preset number can be determined according to actual circumstances.
[0088] This disclosure provides a method for determining target multimedia content that can establish search awareness from multimedia content, and for determining target search terms that can establish search awareness from search terms, based on models and dimensions. This method allows for more flexible determination of target multimedia content and target search terms. At the same time, the target multimedia content and target search terms can better enhance users' understanding of the search engine's search function.
[0089] Regarding the above S103:
[0090] After identifying the target multimedia content and target search terms, they can be pushed out separately through different channels. Specifically, one approach is to first determine the recipients of the target multimedia content and target search terms, and then push them to those recipients through different channels. Alternatively, if there are a large number of target multimedia content and target search terms, they can be further filtered to determine the target multimedia content to be distributed and the target search terms to be distributed, and then pushed out separately through different channels.
[0091] In one alternative implementation, the aforementioned targeted multimedia content and target search terms can be pushed by identifying user groups within specific vertical categories. See also Figure 2The diagram illustrates a method flow for pushing target multimedia content and target search terms through different channels according to an embodiment of this disclosure, including:
[0092] S201, based on the browsing data corresponding to each preset vertical category, determine the distribution object set that matches each preset vertical category; the distribution object set includes multiple target distribution objects.
[0093] S202, for the set of distribution objects, determine the distribution weight corresponding to each preset vertical category; wherein the distribution weight corresponding to the preset vertical category that matches the set of distribution objects is greater than the distribution weight corresponding to other preset vertical categories; the distribution weight is used to indicate the proportion of the distribution quantity of the corresponding preset vertical category in each preset vertical category.
[0094] S203, according to the distribution weight corresponding to each preset vertical category, the target multimedia content and target search terms are distributed to each target distribution object in the distribution object set through different channels.
[0095] Here, the preset vertical categories can include, for example, sports, e-commerce, trending topics, lifestyle services, film and television, food, and so on. During implementation, based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category can be determined; this set of distribution objects includes multiple target distribution objects.
[0096] The browsing data for the preset vertical category can include: the number of times multimedia content within that vertical category has been viewed, the number of times search terms have been triggered, etc. Based on the browsing data, the user group interested in the preset vertical category can be identified, which is the set of distribution targets matched with the preset vertical category. The target distribution targets in the distribution target set are the users interested in the preset vertical category.
[0097] For each set of distribution objects, the distribution weight corresponding to each preset vertical category can be determined; the distribution weight of the preset vertical category matching the distribution object set is greater than the distribution weight of other preset vertical categories. For example, for distribution object set 1, the distribution weight corresponding to the "Hot Topics" category can be determined to be greater than the distribution weights corresponding to the "Life Services" category and the "Film and Television" category. For instance, the distribution weight for the "Hot Topics" category can be determined to be 0.7, the distribution weight for the "Life Services" category to be 0.1, and the distribution weight for the "Film and Television" category to be 0.2. Since the distribution weight can indicate the proportion of distribution quantity of the corresponding preset vertical category within each preset vertical category, the quantity of target multimedia content and target search terms to be distributed under each preset vertical category can be determined according to the distribution weight corresponding to each preset vertical category. The determined quantity of target multimedia content and target search terms can then be distributed to each target distribution object in the distribution object set through different channels.
[0098] Here, by using browsing data corresponding to each preset vertical category, the distribution object set matching each preset vertical category can be determined more accurately. Then, for each distribution object set, the distribution weight corresponding to each preset vertical category is determined. In order to distribute the target multimedia content and target search terms to each target distribution object in the distribution object set through different channels according to the distribution weight corresponding to each preset vertical category, this can better match user needs and further improve users' search awareness of the vertical categories they are interested in.
[0099] For specific implementation, please refer to Figure 3 The diagram illustrates a method flow for determining a set of distribution objects matching each preset vertical category based on browsing data corresponding to each preset vertical category, according to an embodiment of this disclosure. The method includes:
[0100] S2011, for each preset vertical category, determine the same amount of initial distribution information, which includes the target multimedia content for initial distribution and / or the target search terms for initial distribution.
[0101] S2012, distribute the initial distribution information to each distribution object.
[0102] S2013, Based on the browsing data corresponding to each preset vertical category within the preset time period after distribution, determine the set of distribution objects that match each preset vertical category.
[0103] During implementation, the same amount of initial distribution information can be determined for each preset vertical category. The initial distribution information is then distributed to each distribution object. Based on the browsing data corresponding to the initial distribution information in each preset vertical category, the distribution strategy is adjusted to finally determine the set of distribution objects that match each preset vertical category.
[0104] For example, within a preset time period after distribution, such as within 24 hours after distribution, a set of distribution objects matching each preset vertical category is determined based on the browsing data corresponding to each preset vertical category, that is, based on the number of views of the initially distributed target multimedia content under each preset vertical category and / or the number of searches of the initially distributed target search terms. Specifically, for each preset vertical category, based on the number of views of the initially distributed target multimedia content under that preset vertical category by each distribution object and / or the number of searches of the initially distributed target search terms, a set of distribution objects including a third preset number of distribution objects is determined in order of more than less, which is the set of distribution objects matching that preset vertical category. The third preset number can be determined according to actual needs, for example, it can be 100; or it can be determined according to the number of distribution objects and a preset ratio, for example, if the preset ratio is 20% and the number of distribution objects is 1000, then the third preset number is 200.
[0105] Here, by determining the same amount of initial distribution information for each preset vertical category and distributing the initial distribution information to each distribution object, and then based on the browsing data corresponding to each preset vertical category within the preset time period after distribution, the set of distribution objects matching each preset vertical category can be determined more accurately, that is, more accurate push can be achieved, thereby better improving the search recognition of the preset vertical category.
[0106] In another alternative implementation, see Figure 4 The diagram illustrates a method flow for pushing target multimedia content and target search terms through different channels according to an embodiment of this disclosure, including:
[0107] S301, For each preset vertical category, based on the attribute dimensions matched by the preset vertical category, determine the importance score of each target multimedia content and target search term under the preset vertical category.
[0108] S302, based on the distribution status of the preset vertical category on each distribution channel, and the importance scores of each target multimedia content and target search term under the preset vertical category, select the target multimedia content and target search term to be distributed for each preset vertical category.
[0109] S303 will push the target multimedia content and target search terms selected for each preset vertical category to be distributed through different channels.
[0110] Considering the different coverage logics of different vertical categories, for each preset vertical category, the importance score of each target multimedia content and target search term under that preset vertical category can be determined based on the attribute dimensions matched by that preset vertical category. For example, for the hot topic category, where high popularity and authority of information are required, the target multimedia content and target search terms with higher relevance to recent events will have higher importance scores. For the e-commerce category, where more attention needs to be paid to products and brands, the target multimedia content and target search terms with higher matching degree with products and brands will have higher importance scores.
[0111] Furthermore, based on the distribution status of the preset vertical category across various distribution channels, and according to the order of importance scores of the target multimedia content and target search terms under the preset vertical category from high to low, target multimedia content and target search terms to be distributed can be selected for each preset vertical category. Distribution channels can include information feed pages, multimedia content browsing pages, search pages, search results pages, etc. The distribution status can be used to characterize the coverage of the preset vertical category across various distribution channels. This coverage can be, for example, the ratio between the already distributed quantity of target multimedia content and target search terms under the preset vertical category and the distribution quota on each distribution channel. Based on this coverage, the remaining quantity of target multimedia content and target search terms that still need to be distributed to the preset vertical category on each distribution channel can be determined. Therefore, based on the determined remaining quantity, and according to the order of importance scores of the target multimedia content and target search terms under the preset vertical category from high to low, the remaining quantity of target multimedia content and target search terms to be distributed can be selected for each preset vertical category. Subsequently, the target multimedia content and target search terms selected for each preset vertical category can be pushed through different channels.
[0112] In the above method, the importance score of each target multimedia content and target search term under each preset vertical category is determined based on the attribute dimensions matched by each preset vertical category. Based on the distribution status of each preset vertical category on each distribution channel and the importance score of each target multimedia content and target search term under that preset vertical category, the target multimedia content and target search terms to be distributed are selected for each preset vertical category. This can make the coverage of each preset vertical category on each distribution channel more reasonable, improve users' search awareness of the search engine, and better meet users' needs.
[0113] In one optional implementation, the target multimedia content and the target search terms are pushed through different channels, including: displaying the target multimedia content on the information feed recommendation page and displaying the target search terms on the target information page.
[0114] In implementation, target multimedia content and target search terms can be pushed through different channels. Specifically, target multimedia content can be displayed on the information feed push page, or it can be displayed together with multimedia content recommended based on other recommendation rules on the information feed push page. The ranking rule can be to prioritize the display of target multimedia content, and if there are multiple target multimedia content items, to rank them according to their publication time, access popularity, etc. Similarly, target search terms can be displayed on the target information page, or they can be displayed together with search terms recommended based on other recommendation rules on the target information page. The ranking rule can be to prioritize the display of target search terms, and if there are multiple target search terms, to rank them according to their importance, access popularity, etc.
[0115] In practice, the target search term is displayed on the target information page, including: displaying the target search term on at least one of the following pages: the search page where the search was initiated, the search results page, and the multimedia content details page.
[0116] Specifically, on the search page where a search is initiated, the target search term is displayed. This can be shown in the search term area of the search page or in a dropdown menu after the search box is triggered. On the search results page, the target search term can be displayed in a specific search area outside the main search results area, or it can be inserted into the search results. On the multimedia content details page, the target search term can be displayed at the beginning or end of the page, or it can be inserted into the multimedia content details page. By triggering the target search term, users can view the relevant multimedia content associated with that search term, thereby further enhancing their search awareness.
[0117] Those skilled in the art will understand that, in the above-described method of the specific implementation, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.
[0118] Based on the same inventive concept, this disclosure also provides an information push device corresponding to the information push method. Since the principle of the device in this disclosure for solving the problem is similar to the information push method described above in this disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.
[0119] Reference Figure 5The diagram shown is a schematic representation of an information push device according to an embodiment of this disclosure. The device includes: an acquisition module 501, a determination module 502, and a push module 503; wherein,
[0120] The acquisition module 501 is used to acquire multimedia content and search terms from the resource library;
[0121] The determining module 502 is used to determine, from the multimedia content, target multimedia content that can establish search cognition, and from the search terms, target search terms that can establish search cognition;
[0122] The push module 503 is used to push the target multimedia content and the target search terms through different channels respectively.
[0123] In an optional implementation, the push module 503, when pushing the target multimedia content and the target search term through different channels, is used to:
[0124] Based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category is determined; the set of distribution objects includes multiple target distribution objects.
[0125] For the set of distribution objects, a distribution weight is determined for each preset vertical category; wherein the distribution weight of the preset vertical category that matches the set of distribution objects is greater than the distribution weight of other preset vertical categories; the distribution weight is used to indicate the proportion of the distribution quantity of the corresponding preset vertical category in each preset vertical category;
[0126] According to the distribution weights corresponding to each preset vertical category, the target multimedia content and the target search terms are distributed to each target distribution object in the distribution object set through different channels.
[0127] In an optional implementation, the push module 503, when determining the distribution object set matching each preset vertical category based on the browsing data corresponding to each preset vertical category, is used to:
[0128] For each of the preset vertical categories, the same amount of initial distribution information is determined, wherein the initial distribution information includes the target multimedia content to be initially distributed, and / or the target search terms to be initially distributed;
[0129] Distribute the initial distribution information to each distribution object;
[0130] Based on the browsing data corresponding to each preset vertical category within a preset time period after distribution, a set of distribution objects matching each preset vertical category is determined.
[0131] In an optional implementation, the push module 503, when pushing the target multimedia content and the target search term through different channels, is used to:
[0132] For each preset vertical category, based on the attribute dimensions matched by the preset vertical category, the importance scores of each target multimedia content and the target search term under the preset vertical category are determined;
[0133] Based on the distribution status of the preset vertical category on each distribution channel, and the importance scores of each target multimedia content and target search term under the preset vertical category, the target multimedia content and target search term to be distributed are selected for each preset vertical category.
[0134] The target multimedia content and target search terms selected for each of the preset vertical categories will be pushed through different channels.
[0135] In an optional implementation, the determining module 502, when determining target multimedia content capable of establishing search awareness from the multimedia content and target search terms capable of establishing search awareness from the search terms, is configured to:
[0136] Based on multiple search cognitive dimensions, initial multimedia content matching each search cognitive dimension is determined from the multimedia content, and initial search terms matching each search cognitive dimension are determined from the search terms; the search cognitive dimension refers to the category dimension corresponding to the multimedia content that can establish search cognition.
[0137] For each of the search cognitive dimensions, feature clustering is performed on the initial multimedia content and initial search terms to determine at least one first feature condition and at least one second feature condition corresponding to each of the search cognitive dimensions; the first feature condition is matched with the multimedia content, and the second feature condition is matched with the search terms.
[0138] The target multimedia content that meets the first feature condition is determined from the multimedia content, and the target search term that meets the second feature condition is determined from the search term.
[0139] In an optional implementation, the determining module 502, when determining target multimedia content capable of establishing search awareness from the multimedia content and target search terms capable of establishing search awareness from the search terms, is configured to:
[0140] For each multimedia content, first search feature data before browsing the multimedia content, second search feature data after browsing the multimedia content, and first target feature data of not browsing the multimedia content are obtained; and for each search term, third search feature data before initiating a search based on the search term, fourth search feature data after initiating a search based on the search term, and second target feature data of not initiating a search using the search term are obtained.
[0141] The first search feature data, the second search feature data, and the first target feature data corresponding to each of the multimedia contents are input into the trained first search cognitive model to obtain the first search activity improvement value corresponding to each of the multimedia contents. The third search feature data, the fourth search feature data, and the second target feature data corresponding to each of the search terms are input into the trained second search cognitive model to obtain the second search activity improvement value corresponding to each of the search terms.
[0142] Based on the first search activity increase value and the second search activity increase value, the target multimedia content and the target search term are determined from each of the multimedia content and the search term.
[0143] In an optional implementation, the push module 503, when pushing the target multimedia content and the target search term through different channels, is used to:
[0144] The target multimedia content is displayed on the information feed recommendation page, and the target search term is displayed on the target information page.
[0145] In an optional implementation, the push module 503, when displaying the target search term on the target information page, is used for:
[0146] The target search term is displayed on at least one of the following pages: the search page where the search was initiated, the search results page, and the multimedia content details page.
[0147] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.
[0148] This disclosure also provides a computer device, such as... Figure 6 The diagram shown is a schematic representation of a computer device structure provided in an embodiment of this disclosure, including:
[0149] A processor 61 and a memory 62; the memory 62 stores machine-readable instructions executable by the processor 61, and the processor 61 executes the machine-readable instructions stored in the memory 62. When the machine-readable instructions are executed by the processor 61, the processor 61 performs the following steps:
[0150] Retrieve multimedia content and search terms from the resource library;
[0151] From the multimedia content, identify target multimedia content that can establish search awareness, and from the search terms, identify target search terms that can establish search awareness;
[0152] The target multimedia content and the target search terms will be pushed through different channels.
[0153] The aforementioned memory 62 includes a main memory 621 and an external memory 622. The main memory 621, also known as internal memory, is used to temporarily store the computational data in the processor 61, as well as the data exchanged with external memory such as a hard disk. The processor 61 exchanges data with the external memory 622 through the main memory 621.
[0154] The specific execution process of the above instructions can be referred to the steps of the information push method described in the embodiments of this disclosure, and will not be repeated here.
[0155] This disclosure also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the information push method described in the above method embodiments. The storage medium can be a volatile or non-volatile computer-readable storage medium.
[0156] This disclosure also provides a computer program product carrying program code. The program code includes instructions that can be used to execute the steps of the information push method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.
[0157] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.
[0158] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and devices described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. In the several embodiments provided in this disclosure, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection may be through some communication interfaces; the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms.
[0159] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0160] In addition, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0161] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0162] Finally, it should be noted that the above-described embodiments are merely specific implementations of this disclosure, used to illustrate the technical solutions of this disclosure, and not to limit it. The protection scope of this disclosure is not limited thereto. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this disclosure. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be covered within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the protection scope of the claims.
Claims
1. A method for information push, characterized in that, include: Retrieve multimedia content and search terms from the resource library; From the multimedia content, determine the target multimedia content for establishing search cognition, and from the search terms, determine the target search terms for establishing search cognition; The target multimedia content and the target search terms are pushed through different channels respectively; The process of determining target multimedia content for establishing search awareness from the multimedia content and target search terms for establishing search awareness from the search terms includes: Based on multiple search cognitive dimensions, initial multimedia content matching each search cognitive dimension is determined from the multimedia content, and initial search terms matching each search cognitive dimension are determined from the search terms; the search cognitive dimension refers to the category dimension corresponding to the multimedia content used to establish search cognition. For each of the search cognitive dimensions, feature clustering is performed on the initial multimedia content and initial search terms to determine at least one first feature condition and at least one second feature condition corresponding to each search cognitive dimension; the first feature condition is matching the multimedia content, and the second feature condition is matching the search term; and From the multimedia content, determine target multimedia content that meets the first feature condition, and from the search terms, determine target search terms that meet the second feature condition; or, The step of determining target multimedia content for establishing search awareness from the multimedia content, and determining target search terms for establishing search awareness from the search terms, includes: For each multimedia content, first search feature data before browsing the multimedia content, second search feature data after browsing the multimedia content, and first target feature data of not browsing the multimedia content are obtained; and for each search term, third search feature data before initiating a search based on the search term, fourth search feature data after initiating a search based on the search term, and second target feature data of not initiating a search using the search term are obtained. The first search feature data, the second search feature data, and the first target feature data corresponding to each of the multimedia contents are input into a trained first search cognitive model to obtain a first search activity improvement value corresponding to each of the multimedia contents. Similarly, the third search feature data, the fourth search feature data, and the second target feature data corresponding to each of the search terms are input into a trained second search cognitive model to obtain a second search activity improvement value corresponding to each of the search terms. Based on the first search activity increase value and the second search activity increase value, the target multimedia content and the target search term are determined from each of the multimedia content and the search term.
2. The method according to claim 1, characterized in that, The target multimedia content and the target search terms are pushed through different channels, including: Based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category is determined; the set of distribution objects includes multiple target distribution objects. For the set of distribution objects, a distribution weight is determined for each preset vertical category; wherein the distribution weight of the preset vertical category that matches the set of distribution objects is greater than the distribution weight of other preset vertical categories; the distribution weight is used to indicate the proportion of the distribution quantity of the corresponding preset vertical category in each preset vertical category; According to the distribution weights corresponding to each preset vertical category, the target multimedia content and the target search terms are distributed to each target distribution object in the distribution object set through different channels.
3. The method according to claim 2, characterized in that, Based on the browsing data corresponding to each preset vertical category, a set of distribution objects matching each preset vertical category is determined, including: For each of the preset vertical categories, the same amount of initial distribution information is determined, wherein the initial distribution information includes the target multimedia content to be initially distributed, and / or the target search terms to be initially distributed; Distribute the initial distribution information to each distribution object; Based on the browsing data corresponding to each preset vertical category within a preset time period after distribution, a set of distribution objects matching each preset vertical category is determined.
4. The method according to claim 1, characterized in that, The target multimedia content and the target search terms are pushed through different channels, including: For each preset vertical category, based on the attribute dimensions matched by that preset vertical category, the importance scores of each target multimedia content and the target search term under that preset vertical category are determined; Based on the distribution status of the preset vertical category on each distribution channel, and the importance scores of each target multimedia content and target search term under the preset vertical category, the target multimedia content and target search term to be distributed are selected for each preset vertical category. The target multimedia content and target search terms selected for each of the preset vertical categories will be pushed through different channels.
5. The method according to claim 1, characterized in that, The target multimedia content and the target search terms are pushed through different channels, including: The target multimedia content is displayed on the information feed recommendation page, and the target search term is displayed on the target information page.
6. The method according to claim 5, characterized in that, Displaying the target search term on the target information page includes: The target search term is displayed on at least one of the following pages: the search page where the search was initiated, the search results page, and the multimedia content details page.
7. An information push device, characterized in that, include: The acquisition module is used to retrieve multimedia content and search terms from the resource library; The determining module is used to determine, from the multimedia content, target multimedia content for establishing search cognition, and from the search terms, target search terms for establishing search cognition; The push module is used to push the target multimedia content and the target search terms through different channels respectively; The determining module is further configured as follows: Based on multiple search cognitive dimensions, initial multimedia content matching each search cognitive dimension is determined from the multimedia content, and initial search terms matching each search cognitive dimension are determined from the search terms; the search cognitive dimension refers to the category dimension corresponding to the multimedia content used to establish search cognition. For each of the search cognitive dimensions, feature clustering is performed on the initial multimedia content and initial search terms to determine at least one first feature condition and at least one second feature condition corresponding to each search cognitive dimension; the first feature condition is matching the multimedia content, and the second feature condition is matching the search term; and From the multimedia content, determine target multimedia content that meets the first feature condition, and from the search terms, determine target search terms that meet the second feature condition; or, The determining module is further configured as follows: For each multimedia content, first search feature data before browsing the multimedia content, second search feature data after browsing the multimedia content, and first target feature data of not browsing the multimedia content are obtained; and for each search term, third search feature data before initiating a search based on the search term, fourth search feature data after initiating a search based on the search term, and second target feature data of not initiating a search using the search term are obtained. The first search feature data, the second search feature data, and the first target feature data corresponding to each of the multimedia contents are input into a trained first search cognitive model to obtain a first search activity improvement value corresponding to each of the multimedia contents. Similarly, the third search feature data, the fourth search feature data, and the second target feature data corresponding to each of the search terms are input into a trained second search cognitive model to obtain a second search activity improvement value corresponding to each of the search terms. Based on the first search activity increase value and the second search activity increase value, the target multimedia content and the target search term are determined from each of the multimedia content and the search term.
8. A computer device, characterized in that, include: A processor and a memory, the memory storing machine-readable instructions executable by the processor, the processor executing the machine-readable instructions stored in the memory, and when the machine-readable instructions are executed by the processor, the processor performing the steps of the information push method as described in any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a computer device, performs the steps of the information push method as described in any one of claims 1 to 6.
Citation Information
Patent Citations
Hot word recommendation method and device combining multiple data sources
CN106909582A
Advertisement putting method and system
CN107330717A
