Navigation content determination method and deviceTechnical Field
The application relates to the field of internet, in particular to a method and a device for determining navigation content.
Background
At present, in application platforms such as a search engine, a web portal, an e-commerce platform, and the like, in order to facilitate a user to find related content, search results are generally organized according to a certain classification standard, and generally, the results are displayed according to a category form, and the user further screens the category to narrow the range of the search results, thereby helping the user to find target content meeting the search requirement of the user more quickly and accurately. However, due to the limitation of the user knowledge field, the initially input query word sometimes cannot truly and effectively express the real or exact intention of the user, and the user can be prompted to more accurately express the own required intention by the category navigation mode, so that the user is guided to further clarify the own query requirement by clicking navigation information.
In a common e-commerce website, the website establishes its own category system according to different categories of commodity information, and the established category system is generally organized in a tree structure, for example, all commodity information related to a mobile phone is classified into a mobile phone category, and all commodity information related to mp3 is classified into an mp3 category. Before searching for the commodity information, the search engine firstly establishes an index, and marks all the commodity information to be searched with tags respectively, where the tags are used to identify which category or categories corresponding to a plurality of leaf nodes the commodity information belongs to, where the categories corresponding to the leaf nodes are the leaf categories, where the categories corresponding to the leaf nodes are used to represent non-reclassable category units in a tree structure of a category system, for example: for the category tree, the mobile phone is an inseparable category unit, and the mobile phone is the category corresponding to the leaf node of the category tree.
At present, common navigation is generally based on categories corresponding to leaf nodes, and after a query request of a user is received, the categories corresponding to the leaf nodes to which all results returned by search belong are displayed to the user for the user to screen and navigate. For example, when a user searches for a dress, such a search term clearly indicates the search intention of the user, and the category corresponding to the search result corresponding to such a search term is also a category corresponding to a leaf node. In this case, it is reasonable to directly show the categories corresponding to the leaf nodes to the user, however, the user often inputs some broader search terms, such as: the user inputs 'home' as a search term, and for the wider search term, the user may indicate that the user wants to search for various search intentions such as home furniture and bedding, so the obtained search result is also wider and sometimes includes multiple industries or multiple categories.
If all queries determine the navigation content in the manner of the category corresponding to the leaf node, the following problems occur:
first, if a broad query is directly located to a category corresponding to a very accurate leaf node, there may be a case where there is no relationship between the displayed category corresponding to the leaf node and a search result desired by a user, so that the search result is inaccurate, and the user experience is reduced. At this time, the user may re-input other query terms for re-search, and since the client sends a query request to the search engine server many times, a great query pressure may be applied to the search engine server in the case of a large number of search users.
Secondly, if the broad query is located under categories corresponding to a plurality of leaf nodes, the number of recalled search results is generally large, and the results may span the categories corresponding to the plurality of leaf nodes and the categories corresponding to a plurality of non-leaf nodes, and sometimes there may be a plurality of industries in distribution. On one hand, the number of search results is large, and the display results are difficult to cover all intentions of the user under the condition of limited display positions; on the other hand, categories corresponding to leaf nodes of multiple industries are mixed together to be displayed, so that a messy feeling is easily generated for a user, and the user experience is poor. On the other hand, the query matching operation needs to be performed under the categories corresponding to the multiple leaf nodes, so that the data size of the query matching is huge, the efficiency of the search result is reduced, and when the acquired search result with a large number is transmitted to the user client, the network transmission speed is reduced due to the huge data size, and the search efficiency is further influenced.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The application provides a method and a device for determining navigation content, which are used for at least solving the technical problem of low search efficiency caused by directly displaying categories corresponding to all leaf nodes corresponding to search terms in the prior art.
According to an aspect of the present application, there is provided a method of determining navigation content, including: acquiring an input search word; obtaining the click rate of the search result corresponding to the search word in each category on each layer in the category tree according to the historical access record; determining the category to be displayed according to the obtained click rate, displaying the determined category to be displayed on a webpage as navigation content, and displaying the search result under the determined category.
Preferably, determining a category to be displayed according to the obtained click through rate, displaying the determined category to be displayed on a webpage as navigation content, and displaying a search result under the determined category includes: when the click rate of the search result corresponding to the search word in the category corresponding to the first leaf child node in the category tree is larger than or equal to a first preset threshold value, displaying the category corresponding to the first leaf child node as the navigation content; or when the click rate of the search result corresponding to the search term at the category corresponding to each leaf node is not greater than a first predetermined threshold, but the click rate of the category corresponding to one or more leaf nodes is greater than a second predetermined threshold and the categories corresponding to the one or more leaf nodes belong to the same father node, displaying the category corresponding to the leaf node under the father node as the navigation content.
Preferably, determining a category to be displayed according to the obtained click through rate, displaying the determined category to be displayed on a webpage as navigation content, and displaying a search result under the determined category includes: and when the click rate of the search result corresponding to the search word at the category corresponding to each leaf node is smaller than a first preset threshold value, and the click rates of the search result corresponding to the search word at the leaf nodes are larger than a second preset threshold value, the categories corresponding to the leaf nodes do not belong to the same father node, and hierarchical display is performed according to the father-son relationship among the categories.
Preferably, performing hierarchical display according to parent-child relationships among the categories comprises: determining the click rate of a parent node of a category corresponding to each leaf node, wherein the click rate of the parent node is the sum of the click rates of the categories corresponding to all child nodes under the parent node; taking the parent node with the click rate larger than or equal to a third preset threshold value and the categories corresponding to all the child nodes under the parent node with the click rate larger than or equal to the third preset threshold value as the first navigation content to be displayed; and performing hierarchical display on the category corresponding to the parent node and all child nodes under the parent node in the first navigation content to be displayed according to the relationship between the parent node and the child nodes.
Preferably, after determining the click rate of the parent node of the category corresponding to each leaf node, the method further includes: determining the click rate of an upper node of the parent node which is smaller than the third preset threshold, wherein the click rate of the upper node is the sum of the click rates of all categories corresponding to all child nodes under the upper node; taking the category corresponding to the upper-layer node with the click rate larger than or equal to a fourth preset threshold and all child nodes under the upper-layer node with the click rate larger than or equal to the fourth preset threshold as second navigation content to be displayed; and performing hierarchical display on the first navigation content to be displayed and the second navigation content to be displayed according to a parent-child relationship between the first navigation content to be displayed and the second navigation content to be displayed.
Preferably, the historical access record comprises: the click rate of the search result corresponding to the search term and/or the click rate of each category in the navigation content corresponding to the search term.
Preferably, the obtaining of the click rate of the search result corresponding to the search term in each category on each layer in the category tree according to the historical access record includes: setting weight values for all categories according to the historical access records, wherein the categories with more clicks correspond to large weight values, and the categories with less clicks correspond to small weight values; and determining the click rate of each category according to the weight value set for each category.
According to another aspect of the present application, there is provided a navigation content presentation apparatus including: a first acquisition unit configured to acquire an input search word; the second acquisition unit is used for acquiring the click rate of the search result corresponding to the search word in each category on each layer in the category tree according to the historical access record; and the determining unit is used for determining the category to be displayed according to the acquired click rate, displaying the determined category to be displayed on the webpage as the navigation content, and displaying the search result under the determined category.
Preferably, the determining unit includes: a first display module, configured to display a category corresponding to a first leaf node in the category tree as the navigation content when a click rate of a category corresponding to the first leaf node of a search result corresponding to the search term in the category tree is greater than or equal to a first predetermined threshold; or the second display module is configured to display the category corresponding to the leaf node under the parent node as the navigation content when the click rate of the search result corresponding to the search term at each leaf node is not greater than the first predetermined threshold, but the click rate of the category corresponding to one or more leaf nodes is greater than the second predetermined threshold and the categories corresponding to the one or more leaf nodes belong to the same parent node.
Preferably, the determining unit includes: and the third display module is used for carrying out hierarchical display according to the parent-child relationship among the categories when the click rate of the search result corresponding to the search word at the category corresponding to each leaf node is smaller than a first preset threshold value and the click rate of the search result corresponding to the search word at the category corresponding to each leaf node is larger than a second preset threshold value, wherein the categories corresponding to a plurality of leaf nodes do not belong to the same parent node.
According to the method and the device, the click condition of the related category of the current search word is obtained by analyzing the historical click behavior of the user, so that different navigation contents are determined to be displayed according to different conditions, the displayed navigation contents are more accurate, the user is prevented from searching again due to the fact that the displayed navigation contents are inaccurate, and the query pressure on a search engine server caused by searching again is reduced; meanwhile, as the more accurate navigation content (the category corresponding to the leaf node) is returned instead of the categories corresponding to all the leaf nodes corresponding to the search word, the number of the search results under the returned navigation content (the category corresponding to the leaf node) is reduced under the condition of ensuring the accuracy, the number of the search results transmitted by the network is reduced, and the return speed and the search efficiency of the search results are improved. By means of the method, the technical problem that in the prior art, the searching efficiency is low due to the fact that the categories corresponding to all the leaf nodes corresponding to the searching words are directly displayed is effectively solved, and the technical effect of improving the searching efficiency is achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 is a block diagram of a preferred architecture of a navigation content determination system according to an embodiment of the present application;
fig. 2 is a block diagram of a preferred structure of a navigation content determining apparatus according to an embodiment of the present application;
FIG. 3 is a preferred diagram of a web page to be presented with navigation content according to an embodiment of the present application;
fig. 4 is a block diagram of another preferred structure of a navigation content determining apparatus according to an embodiment of the present application;
FIG. 5 is a preferred diagram of a category tree according to an embodiment of the present application;
FIG. 6 is another preferred diagram of a category tree according to an embodiment of the present application;
FIG. 7 is a preferred flow chart of a method of determining navigation content according to an embodiment of the present application;
FIG. 8 is another preferred flow chart of a method of determining navigation content according to an embodiment of the present application;
FIG. 9 is a schematic diagram of a preferred architecture of a navigation content determination system according to an embodiment of the present application;
fig. 10 is another preferred flowchart of a navigation content determining method according to an embodiment of the present application.
Detailed Description
The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Before describing further details of embodiments of the present application, one suitable computing architecture that may be used to implement the principles of the present application will be described with reference to FIG. 1. In the following description, embodiments of the present application will be described with reference to acts and symbolic representations of operations that are performed by one or more computers, unless indicated otherwise. As such, it will be understood that such acts and operations, which are at times referred to as being computer-executed, include the manipulation by the processing unit of the computer of electrical signals representing data in a structured form. This manipulation transforms the data or maintains it at locations in the computer's memory system, which reconfigures or otherwise alters the operation of the computer in a manner well understood by those skilled in the art. The data structures that maintain the data are physical locations of the memory that have particular properties defined by the format of the data. However, while the present application is described in the foregoing context, it is not meant to be limiting, as those of skill in the art will appreciate that aspects of the acts and operations described hereinafter may also be implemented in hardware.
Turning to the drawings, wherein like reference numerals refer to like elements, the principles of the present application are illustrated as being implemented in a suitable computing environment. The following description is based on illustrated embodiments of the application and should not be taken as limiting the application with regard to alternative embodiments that are not explicitly described herein.
FIG. 1 shows a schematic diagram of one example computer architecture that may be used for these devices. For descriptive purposes, the architecture portrayed is only one example of a suitable environment and is not intended to suggest any limitation as to the scope of use or functionality of the application. Neither should the computing system be interpreted as having any dependency or requirement relating to any one or combination of components illustrated in FIG. 1.
The principles of the present application may be implemented using other general purpose or special purpose computing or communication environments or configurations. Examples of well known computing systems, environments, and configurations that may be suitable for use with the application include, but are not limited to, personal computers, servers, multiprocessor systems, microprocessor-based systems, minicomputers, mainframe computers, and distributed computing environments that include any of the above systems or devices.
In its most basic configuration, the navigation content determination system 100 of fig. 1 includes at least: a server 102 of a web site and one or more clients 104. Server 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA, a storage device for storing data, and a transmission device for communicating with clients; the client 104 may include: the system comprises a microprocessor MCU, a transmission device communicated with a server and a display device interacted with a user. In this description and in the claims, a "system" may also be defined as any hardware component or combination of hardware components capable of executing software, firmware, or microcode to perform a function, or the presentation system 100 may even be distributed to perform distributed functions.
As used herein, the terms "module," "component," or "unit" may refer to a software object or routine that executes on the presentation system 100 for navigating content. The different components, modules, units, engines, and services described herein may be implemented as objects or processes that execute (e.g., as separate threads) on the presentation system 100 navigating content. Although the systems and methods described herein are preferably implemented in software, implementations in hardware or a combination of software and hardware are also possible and contemplated.
Example 1
Fig. 2 is a schematic diagram of a preferred structure of a device 200 for determining navigation content according to an embodiment of the present application, which is preferably located on the server 102 or the client 104.
In the embodiment of the present application, the displayed web page mainly consists of two parts, as shown in fig. 3, including: the navigation system comprises a navigation area and a search result area, wherein the navigation area displays related navigation information and is used for displaying clickable category classification to a user, so that the user can search more conveniently; the search result area is used for displaying a search result corresponding to a search word input by the user or displaying a search result triggered after the user clicks a certain category in the navigation information.
Preferably, the navigation content determining apparatus 200 shown in fig. 2 may include: a first acquisition unit 202, a second acquisition unit 204 and a determination unit 206. These several structures are described in detail below.
A first acquisition unit 202 for acquiring an input search term;
a second obtaining unit 204, coupled to the first obtaining unit 202, configured to obtain, according to the historical access record, click rates of search results corresponding to respective categories on each layer in the category tree;
for example, the user inputs the search term "home", and accordingly obtains a historical access record of a corresponding search result when the search term "home" is used as the search term, and preferably, the historical access record includes: the click rate of the search result corresponding to the search word and/or the click rate of each category in the navigation content corresponding to the search word. In the preferred embodiment, the historical visit records of the user refer to the historical visit records of all users stored in the website database based on the search word and do not refer to the historical visit records of the user alone. Meanwhile, the historical access records may be the click-through rate of the user corresponding to each specific search result, or the click-through rate corresponding to each category in the navigation content corresponding to "home". For example, when searching for "home", the click rate of the user on "sofa" is 30%, the click rate of the user on "table" is 20%, and so on, then these data counted by the historical access records are the click rate corresponding to the search word that needs to be obtained. The navigation can be used for conducting corresponding guidance on searching or browsing of the user under a certain category system, the category is a classification system related to texts or product topics, and the navigation can play a role in navigation in searching and assist the user in searching.
The determining unit 206 is coupled to the second obtaining unit 204, and configured to determine a category to be displayed according to the obtained click rate, display the determined category to be displayed on a webpage as navigation content, and display a search result under the determined category.
In the embodiment, the click condition of the related category of the current search term is obtained by analyzing the historical click behavior of the user, so that different navigation contents are determined to be displayed according to different conditions, the displayed navigation contents are more accurate, the user is prevented from re-searching due to inaccurate displayed navigation contents, and the query pressure on a search engine server caused by re-searching is reduced; meanwhile, as the more accurate navigation content (the category corresponding to the leaf node) is returned instead of the categories corresponding to all the leaf nodes corresponding to the search word, the number of the search results under the returned navigation content (the category corresponding to the leaf node) is reduced under the condition of ensuring the accuracy, the number of the search results transmitted by the network is reduced, and the return speed and the search efficiency of the search results are improved. By means of the method, the technical problem that in the prior art, the searching efficiency is low due to the fact that the categories corresponding to all the leaf nodes corresponding to the searching words are directly displayed is effectively solved, and the technical effect of improving the searching efficiency is achieved.
After the click rate of the search result corresponding to the search term in each category is obtained, the specific navigation display mode of the search term can be determined based on different click rate distributions. In a preferred embodiment, as shown in fig. 4, the determining unit 206 may include: a first display module 402 and a second display module 404.
The first display module 402 is configured to display the category corresponding to the first leaf node as the navigation content when the click rate of the search result corresponding to the search term in the category corresponding to the first leaf node is greater than or equal to a first predetermined threshold; for example, if the click rate of a leaf node in the history access record corresponding to a certain search word is 99%, it can be determined that the search requirement of the user can be met under the category corresponding to the leaf node, so that the category corresponding to the leaf node and the search result corresponding to the category corresponding to the leaf node can be directly displayed, and the user can obtain an object to be searched by himself through one search, which is called direct category transfer. The category corresponding to the leaf node is the smallest and non-separable category unit in the category tree, such as the more specific search terms like "one-piece dress".
However, the search results corresponding to all the search terms are not concentrated in the category corresponding to one leaf node, or may be concentrated in the categories corresponding to a plurality of leaf sub-nodes corresponding to the category corresponding to a certain non-leaf node, for this case, the category focusing may be performed, and the corresponding navigation display manner may be to display the categories corresponding to all the leaf nodes under the category corresponding to the non-leaf node, or to display only the categories corresponding to a few more concentrated leaf nodes under the category corresponding to the non-leaf node as navigation contents, where the concentration is a relative concept, and the concentration degree of the click may be determined according to the historical click behavior of each node, and preferably, the concentration degree here refers to the click rate of the node in the category tree hierarchy where the node is located after counting one click rate of all the users, then, whether the number of the categories is larger than a preset threshold value is determined, so that more concentrated categories in the categories are determined. In a preferred embodiment, the second presentation module 404 is configured to present all leaf nodes corresponding to a parent node as navigation content when the click through rate of the search result corresponding to the search term at each leaf node is not greater than the first predetermined threshold, but the click through rate of the category corresponding to one or more leaf nodes is greater than the second predetermined threshold and the categories corresponding to the one or more leaf nodes belong to the same parent node. The following description will be made in detail with reference to a specific embodiment:
as shown in fig. 5, which is a preferred diagram of a category tree, as can be seen from the content shown in fig. 5, among the categories corresponding to all leaf nodes, only B21 and B23 are concentrated, and all the concentrated nodes are found out as the nodes finally shown. For example, if the click rate of the category corresponding to the leaf node is greater than 30%, the category (also referred to as a node) in the layer is more concentrated. Meanwhile, it can be found that both B21 and B23 are categories corresponding to leaf nodes of B2, and therefore, it is conceivable that the click demand of the user is almost satisfied under B2. Thus, the categories corresponding to the direct child nodes of B2 may be: b21, B22, and B23 are displayed as navigation content to be displayed. Preferably, only B21 and B23 may also be displayed or B2 may also be presented as navigation content, which is so-called category focusing.
In practical application, a situation that categories corresponding to a plurality of leaf nodes with higher click rates in a history access record corresponding to a certain search word do not belong to the same category corresponding to a non-leaf node may also occur, so that contents to be displayed in a manner of focusing the categories are more, and in this situation, hierarchical display can be performed according to a parent-child relationship between the categories. That is, it is determined according to the click rate of each category which non-leaf nodes correspond to which categories are displayed, and which categories correspond to the leaf nodes corresponding to the categories, and this display mode may be called a category hierarchy. In a preferred embodiment, the determining unit comprises: and the third display module is used for carrying out hierarchical display according to the parent-child relationship among the categories when the click rate of the search result corresponding to the search word at each leaf node is smaller than the first preset threshold value, and the categories corresponding to the leaf nodes of which the click rate is larger than the second preset threshold value do not belong to the same parent node.
Specifically, when the category hierarchy is displayed, it may be determined whether the click rate of the parent node of the category corresponding to the leaf node is relatively concentrated, and preferably, if the click rate of the parent node is not concentrated, the parent node may continue to be pushed to the upper layer until the parent node or the grandparent of the category corresponding to the leaf node belongs to the more concentrated node in the layer corresponding to the parent node or the grandparent of the category corresponding to the leaf node or the grandparent is pushed to the root node. That is, if the category corresponding to the leaf node corresponding to a search term is relatively dispersed in the lowest layer corresponding to the leaf node, that is, the click rate corresponding to the leaf node is relatively low, it is determined whether the click rate of the parent node is relatively concentrated in the layer corresponding to the parent node, that is, whether the click rate corresponding to the parent node is relatively high, if yes, the categories corresponding to all child nodes under the parent node corresponding to the parent node are displayed as final navigation content, and if the click rate of the parent node corresponding to the leaf category is not high, the parent node continues to perform upper-layer judgment until a node belonging to relatively concentrated in the corresponding layer or a root node is found. Preferably, the click rate of each upper node is obtained by performing an accumulation operation on the click rate of the corresponding lower node. The following description will be made of a specific example of the category hierarchy presentation:
fig. 6 is another preferred diagram of the category tree, where the category hierarchy is searched upward from the category corresponding to the leaf node, and the search determination is not stopped until the corresponding parent node or ancestor node is found to be centralized in the corresponding layer. In the category tree shown in fig. 6, since B12 is distributed at its corresponding level, it is determined whether its parent node (B1) is centralized, and if so, the categories (B11, B12) corresponding to B1 and all its corresponding child nodes are all presented as navigation contents, and similarly, for C11 and C12, themselves are distributed, their corresponding parent nodes (C1) are also distributed, but their corresponding ancestor nodes (C) are centralized, and therefore, the categories (C11 and C12) corresponding to its ancestor node C and its child nodes are presented as navigation contents. Therefore, for the category tree shown in fig. 6, what is finally displayed hierarchically as the navigation content is: c: c1, C2; b1: b11, B12; b2: b21, B22, B23; e3: e31, E32. The hierarchical display mode may be a display mode as shown in fig. 3, that is, hierarchical display is performed according to a parent-child relationship between each other.
In the above preferred embodiments, the historical access records include, but are not limited to, click-through rates of search results corresponding to search terms, and/or click-through rates of various categories in navigation content corresponding to search terms.
In order to better confirm the click rate of each category node in the category structure, a weighted value can be set for each category, the correspondingly set weighted value can be determined according to the click condition in the historical access record, a large weight can be set when the click rate is high, and a small weight can be set when the click rate is low. In a preferred embodiment, the obtaining the click rate of the search result corresponding to the search term in each category according to the historical access record of the user includes: setting weight values for all categories according to historical access records, wherein the categories with a large number of clicks correspond to large weight values, and the categories with a small number of clicks correspond to small weight values; and determining the click rate of each category according to the weight value set for each category.
Example 2
In the embodiment of the present application, the displayed web page mainly consists of two parts, as shown in fig. 3, including: the navigation system comprises a navigation area and a search result area, wherein the navigation area displays related navigation information and is used for displaying clickable category classification to a user, so that the user can search more conveniently; the search result area is used for displaying a search result corresponding to a search word input by the user or displaying a search result triggered after the user clicks a certain category in the navigation information.
On the basis of the devices shown in fig. 2 and 4, the application provides a preferred method for determining navigation content. As shown in fig. 7, the method may include the steps of:
step S702: acquiring an input search word;
step S704: acquiring the click rate of the search result corresponding to the search word in each category on each layer in the category tree according to the historical access record;
for example, the user inputs the search term "home", and accordingly obtains a historical access record of a corresponding search result when the search term "home" is used as the search term, and preferably, the historical access record includes: the click rate of the search result corresponding to the search word and/or the click rate of each category in the navigation content corresponding to the search word. In the preferred embodiment, the historical visit records of the user refer to the historical visit records of all users stored in the website database based on the search word and do not refer to the historical visit records of the user alone. Meanwhile, the historical access records may be the click-through rate of the user corresponding to each specific search result, or the click-through rate corresponding to each category in the navigation content corresponding to "home". For example, when searching for "home", the click rate of all users for "sofa" is 30%, the click rate for "table" is 20%, and so on, then these data counted by the historical access records are the click rate corresponding to the search word that needs to be obtained. The navigation can be used for conducting corresponding guidance on searching or browsing of the user under a certain category system, the category is a classification system related to texts or product topics, and the navigation can play a role in navigation in searching and assist the user in searching.
Step S706: determining the category to be displayed according to the obtained click rate, displaying the determined category to be displayed on a webpage as navigation content, and displaying the search result under the determined category.
In the embodiment, the click condition of the related category of the current search term is obtained by analyzing the historical click behavior of the user, so that different navigation contents are determined to be displayed according to different conditions, the displayed navigation contents are more accurate, the user is prevented from re-searching due to inaccurate displayed navigation contents, and the query pressure on a search engine server caused by re-searching is reduced; meanwhile, as the more accurate navigation content (the category corresponding to the leaf node) is returned instead of the categories corresponding to all the leaf nodes corresponding to the search word, the number of the search results under the returned navigation content (the category corresponding to the leaf node) is reduced under the condition of ensuring the accuracy, the number of the search results transmitted by the network is reduced, and the return speed and the search efficiency of the search results are improved. By means of the method, the technical problem that in the prior art, the searching efficiency is low due to the fact that the categories corresponding to all the leaf nodes corresponding to the searching words are directly displayed is effectively solved, and the technical effect of improving the searching efficiency is achieved.
After the click rate of the search result corresponding to the search term in each category is obtained, the specific navigation display mode of the search term can be determined based on different click rate distributions. Preferably, the display of the navigation content can be performed in the following ways:
when the click rate of the search result corresponding to the search word on the category corresponding to the first leaf child node is larger than or equal to a first preset threshold value, displaying the category corresponding to the first leaf child node as navigation content; that is, only categories corresponding to leaf nodes having the click through rate greater than or equal to the first predetermined threshold may be presented in the navigation area at this time, and the search results corresponding to the search term may be directly presented in the result presentation area. For example, if the click rate of a leaf node in the history access record corresponding to a certain search word is 99%, it can be determined that the search requirement of the user can be met under the category corresponding to the leaf node, so that the category corresponding to the leaf node and the search result corresponding to the category corresponding to the leaf node can be directly displayed, and the user can obtain an object to be searched by himself through one search, which is called direct category transfer. The category corresponding to the leaf node is the smallest and non-separable category unit in the category tree, such as the more specific search terms like "one-piece dress".
However, the search results corresponding to all the search terms are not concentrated in the category corresponding to one leaf node, or may be concentrated in the categories corresponding to a plurality of leaf nodes corresponding to a category corresponding to a certain non-leaf node, and in this case, the category focusing may be performed, and the corresponding navigation display manner may be to display the categories corresponding to all the leaf nodes under the category corresponding to the non-leaf node, or to display only the categories corresponding to a few leaf nodes with higher click rates under the category corresponding to the non-leaf node as navigation contents, where the concentration is a relative concept, and the concentration degree of the click may be determined according to the historical click behavior of each node, and preferably, the concentration degree here refers to the click rate of the node in the category tree hierarchy where the node is located after counting one click rate of all the users, then, whether the number of the categories is larger than a preset threshold value is determined, so that more concentrated categories in the categories are determined. In a preferred embodiment, when the click rate of the search result corresponding to the search term at the category corresponding to each leaf node is not greater than the first predetermined threshold, but the click rate of the category corresponding to one or more leaf nodes is greater than the second predetermined threshold and the categories corresponding to one or more leaf nodes belong to the same parent node, the categories corresponding to all the leaf nodes under the parent node are displayed as the navigation content.
The following description will be made in detail with reference to a specific embodiment:
as shown in fig. 5, which is a preferred diagram of a category tree, as can be seen from the content shown in fig. 5, among the categories corresponding to all leaf nodes, only B21 and B23 are concentrated, and all the concentrated nodes are found out as the nodes finally shown. For example, if the click rate of the category corresponding to the leaf node is greater than 30%, the category (also referred to as a node) in the layer is more concentrated. Meanwhile, it can be found that both B21 and B23 are categories corresponding to leaf nodes of B2, and therefore, it is conceivable that the click demand of the user is almost satisfied under B2. Thus, the categories corresponding to the direct child nodes of B2 may be: b21, B22, and B23 are displayed as navigation content to be displayed. Preferably, only B21 and B23 may also be displayed or B2 may also be presented as navigation content. This is called category focusing.
In practical application, a situation that categories corresponding to a plurality of leaf nodes with higher click rates in a history access record corresponding to a certain search word do not belong to the same category corresponding to a non-leaf node may also occur, so that contents to be displayed in a manner of focusing the categories are more, and in this situation, hierarchical display can be performed according to a parent-child relationship between the categories. That is, it is determined according to the click rate of each category which non-leaf nodes correspond to which categories are displayed, and which categories correspond to the leaf nodes corresponding to the categories, and this display mode may be called a category hierarchy. In a preferred embodiment, the obtaining of the recall rate of the search result corresponding to the search term in each category according to the historical access record of the user includes: and when the click rate of the search result corresponding to the search word at the category corresponding to each leaf node is smaller than a first preset threshold value, and a plurality of leaf categories of which the click rate is larger than a second preset threshold value do not belong to the same father node, performing hierarchical display according to the father-son relationship among the categories.
Specifically, when performing category hierarchy presentation, the method may be implemented as shown in fig. 8, and includes the following steps:
step S802: determining the click rate of a parent node of the category corresponding to each leaf node, wherein the click rate of the parent node is the sum of the click rates of the categories corresponding to all child nodes under the parent node;
step S804: taking the father node with the click rate larger than or equal to a third preset threshold value and the categories corresponding to all the child nodes under the father node as the first navigation content to be displayed;
step S806: determining the click rate of an upper node of the father node which is smaller than a third preset threshold value, wherein the click rate of the upper node is the sum of the click rates of all child nodes under the upper node;
step S808: taking the upper node with the click rate larger than or equal to a fourth preset threshold value and all the sub-nodes under the upper node as second navigation contents to be displayed, namely, taking the determined upper node and the categories corresponding to the direct sub-nodes as the navigation contents to be displayed; preferably, the third threshold and the fourth threshold for the above may be values determined according to the characteristics of each hierarchy, and there is no inevitable relation between the first and second thresholds.
Step S810: and carrying out hierarchical display on the first navigation content to be displayed and the second navigation content to be displayed according to the parent-child relationship between the first navigation content to be displayed and the second navigation content to be displayed.
That is, it is determined whether the click rates of the parent nodes of the categories corresponding to the leaf nodes are relatively centralized, and preferably, if the click rates of the parent nodes are not centralized, the click rates can be continuously pushed to the upper layer until the parent node or the grandparent node of the category corresponding to the leaf node belongs to the more centralized nodes in the layer corresponding to the parent node or the grandparent node is pushed to the root node. That is, if the category corresponding to the leaf node corresponding to a search term is relatively dispersed in the corresponding lowest layer, that is, the click rate corresponding to the leaf node is relatively low, it is determined whether the click rate of the parent node is relatively concentrated in the layer corresponding to the parent node, that is, whether the click rate corresponding to the parent node is relatively high, if so, all child nodes under the parent node corresponding to the parent node are displayed as final navigation content, and if the click rate of the parent node corresponding to the leaf node is not high, the parent node continues to perform upper-layer judgment until a node belonging to relatively concentrated in the corresponding layer or a root node is found. Preferably, the click rate of each upper node is obtained by performing an accumulation operation on the click rate of the corresponding lower node. The following description will be made of a specific example of the category hierarchy presentation:
fig. 6 is another preferred diagram of the category tree, where the category hierarchy is searched upward from the category corresponding to the leaf node, and the search determination is not stopped until the corresponding parent node or ancestor node is found to be centralized in the corresponding layer. In the category tree shown in fig. 6, since B12 is distributed at its corresponding level, it is determined whether its parent node (B1) is centralized, and if so, the categories (B11, B12) corresponding to B1 and all its corresponding child nodes are all presented as navigation contents, and similarly, for C11 and C12, themselves are distributed, their corresponding parent nodes (C1) are also distributed, but their corresponding ancestor nodes (C) are centralized, and therefore, the categories (C11 and C12) corresponding to its ancestor node C and its child nodes are presented as navigation contents. Therefore, for the category tree shown in fig. 6, what is finally displayed hierarchically as the navigation content is: c: c1, C2; b1: b11, B12; b2: b21, B22, B23; e3: e31, E32. The hierarchical display mode may be a display mode as shown in fig. 3, that is, hierarchical display is performed according to a parent-child relationship between each other.
In the above preferred embodiments, the historical access records include, but are not limited to, click-through rates of search results corresponding to search terms, and/or click-through rates of various categories in navigation content corresponding to search terms.
In order to better confirm the click rate of each category node in the category structure, a weighted value can be set for each category, the correspondingly set weighted value can be determined according to the click condition in the historical access record, a large weight can be set when the click rate is high, and a small weight can be set when the click rate is low. In a preferred embodiment, the obtaining the click rate of the search result corresponding to the search term in each category according to the historical access record of the user includes: setting weight values for all categories according to historical access records, wherein the categories with a large number of clicks correspond to large weight values, and the categories with a small number of clicks correspond to small weight values; and determining the click rate of each category according to the weight value set for each category.
Example 3
The present application provides a preferred embodiment to further explain the present application, but it should be noted that the preferred embodiment is only for better describing the present application and should not be construed as unduly limiting the present application.
Aiming at the following defects in the prior art: firstly, completely displaying categories corresponding to leaf nodes to give a user a single navigation form; secondly, the navigation content of the wide query is not friendly, the navigation category display is disordered, and the user is not facilitated to find the set of the required result categories. It is difficult to accurately determine the actual intent of the user search by directly navigating the categories corresponding to the leaf nodes, and sometimes erroneous positioning is given.
In the method provided by the application, the distribution of click rates of various categories in the search results can be determined through analysis of the search terms of the user and according to the historical access records of the user, and different navigation forms are given according to different requirements, that is, navigation based on different granularities is given to broad and accurate queries, and preferably, the following modes can be included: through categories, focused categories, hierarchical categories, and tiled categories, etc., thereby effectively determining the user's search intent in order to more accurately determine search results. Wherein,
the category is through, that is, the search terms of the user are directly positioned to a certain most relevant category, so that the user can obtain the result of the desired search by screening once.
The category focusing is to directly display several most relevant categories aiming at the search terms and automatically shield other non-relevant categories so as to narrow the screening range of the user and improve the screening efficiency.
The category hierarchy is to show different sub-trees according to a category tree structure based on search words given by a user, give a path of category parent-child relation, and help the user to determine a search intention at a higher speed when the user performs broad search.
The category is spread, that is, when the search intention of the user is particularly wide, the categories corresponding to all leaf nodes related to the search terms are directly displayed.
The following is a detailed description of several of the above approaches:
determining the click condition of the category corresponding to each leaf node according to the distribution of the click rate in the historical access records so as to provide different navigation forms:
1) for broad queries, the search structure of the category includes categories corresponding to a plurality of non-leaf nodes, and in this case, the result can be displayed hierarchically according to the architecture of the category, which is called category hierarchy and also called hierarchy navigation;
2) for more accurate and specific query, the categories corresponding to the leaf nodes to which the search results belong are relatively concentrated, and generally concentrated in the categories corresponding to the leaf nodes under the categories corresponding to a certain non-leaf node, so that the categories can be directly positioned to the leaf nodes under the non-leaf nodes for the user to screen, and the category focusing is called;
3) for very accurate and specific search terms, when most of the corresponding search results belong to the category corresponding to one leaf node, the search results can be directly positioned to the category corresponding to the one leaf node, and the results of the categories corresponding to other leaf nodes are automatically filtered by a search engine, so that the search results are called as category direct;
4) if the distribution of the search structure can be covered by a plurality of leaf nodes (i.e. the categories corresponding to the above leaf nodes), but can not be specific to some specific upper-layer nodes, several leaf nodes covering the search result can be directly displayed as navigation content, and the manner becomes category tiling.
Namely, according to different query intentions, different navigation forms are given to a user, and specifically, the method includes four types: category through, category focus, category hierarchy, and category tiling.
In the mode, firstly, the relevance between the original query of the user and the selected category is analyzed according to the relevance between texts and large-scale historical user behaviors, so that the query intention of the user is positioned to the categories corresponding to all relevant leaf nodes, then, a specific navigation display form is judged, on one hand, the intention range of a query word is analyzed, namely, whether the query intention of the user is wide or specific is determined, on the other hand, the categories corresponding to all relevant leaf nodes are constructed into a category tree structure corresponding to the category tree structure by means of a category architecture, on the other hand, all tree nodes are endowed with different weights according to the clicking condition, and finally, the displayed hierarchy is judged through threshold value analysis.
The embodiment of the present application further provides an overall architecture of a system for implementing the above search method, as shown in fig. 9, the system includes four parts:
the front-end server is used for receiving the query of a user and distributing a request to a smart query platform (a platform containing a navigation algorithm) and a search engine;
the smart query platform is an algorithm platform for processing queries, and the navigation display method service can be deployed on a machine of the platform;
a search engine for querying results;
and the operation tool server is used for carrying out manual operation.
Based on the above system architecture, the embodiment of the present application further provides a preferred processing flow, as shown in fig. 10, including (step S1002 to step S1012):
step S1002: and acquiring the search terms requested by the front end of the search engine.
Step S1004: and judging the category tendency of the search word by analyzing historical access records in a large-scale user historical log, wherein the category in the step refers to a category corresponding to a leaf node in a category structure.
Step S1006: judging whether the query is direct, preferably, the direct category is a query word which is very clear for the user to search, analyzing almost all (99%) of user click behaviors through the historical behaviors of the website and concentrating on the category, namely, the condition that the user's requirements can be basically met in the category, displaying the navigation in a direct category mode, and finally displaying the result that although the user only searches for the query word once, the actually obtained result is the result of the result obtained by the search word after being filtered by the category, so that the result obtained by the user query is more accurate, the path of the user search is effectively shortened, and the experience effect of the user is improved.
Step S1008: judging whether the category is focused, wherein the category is focused means that for the search terms of the user, the returned results of the search engine may be distributed in a plurality of categories, but the historical click behaviors of the user are generally concentrated in a certain number of categories (for example, 3 categories), that is, the requirements of the user can be basically and completely met under the categories. In this case, the navigation is presented in a category focused manner. The effect finally shown to the user is that the result returned by a certain search word of the user is distributed in a plurality of categories, however, navigation only shows a certain number of the categories, the screening range of the user can be effectively reduced by the method, the user is helped to automatically shield other non-related categories, and the effects of reducing screening cost and improving user experience are achieved.
Step S1010: judging whether the search term is a category hierarchy, wherein the category hierarchy is generally used under the conditions that the search terms of a user are wide, the returned search results and the historical click behaviors of the user are relatively dispersed, on one hand, the categories corresponding to leaf nodes are distributed in a large number, and on the other hand, a plurality of parent nodes of the categories corresponding to the leaf nodes are distributed on a plurality of different categories, so that the search intention of the user is difficult to be directly positioned to the categories corresponding to some specific leaf nodes under the search condition, if the categories corresponding to all the leaf nodes are directly displayed to the user, the number of the categories corresponding to the leaf nodes to be displayed is too large and complicated, and the categories are not organized, so that the user is difficult to find the required category navigation with low cost, the user experience is poor, and the click behaviors of the user in a navigation area are influenced, under the condition, the categories corresponding to all the leaf nodes can be piled according to a hierarchical navigation mode, and a proper category sub-tree is extracted from the category tree according to the user click probability and the like to serve as a category classification method.
That is, as shown in fig. 6, the description is given by taking a category tree as a three-level category as an example, a category tree is constructed according to a category architecture, and the concentration degree of clicks is determined according to the click behavior of each category. In the category hierarchy, the searching and the judging are carried out in a manner of upwards searching from the category corresponding to the leaf node, and the searching and the judging are stopped until the corresponding father node or ancestor node is found to be concentrated in the corresponding layer.
In the category tree shown in fig. 6, since B12 is distributed at its corresponding level, it is determined whether its parent node (B1) is centralized, and in the case of yes determination, B1 and all its corresponding child nodes are presented as navigation contents, and similarly, for C11 and C12, themselves are distributed, and its corresponding parent node (C1) is also distributed, whereas its corresponding ancestor node (C) is centralized, and therefore, its ancestor node C and its child nodes (C11 and C12) are presented as navigation contents. Thus, for fig. 6, what is finally shown is: c: c1, C2; b1: b11, B12; b2: b21, B22, B23: e3: e31, E32.
Step S1012: if all the scenes are not met, the search result corresponding to the search word can be covered in the categories corresponding to a certain number of leaf nodes, but cannot be reasonably distributed to a certain scene, and in this case, a category tiled display mode can be adopted, that is, the categories corresponding to all the leaf nodes which can be covered by the category tiled display mode are displayed as final navigation content.
In the embodiment, the categories corresponding to all related leaf nodes of the current search word are obtained by analyzing the historical clicking behaviors of the user, then the judgment is carried out according to different scenes according to the sequence of direct category, focusing category, hierarchy of category and tiling category, and the exit is carried out when a certain scene which is in accordance with the judgment is determined, and the scene which is in accordance with the judgment is taken as the final output result of the navigation content.
As can be seen from the above description, the present application has the following advantages: the method has the advantages that the historical clicking behavior of the user is analyzed to obtain the clicking condition of the related category of the current search word, so that different navigation contents are determined to be displayed according to different conditions, the displayed navigation contents are more accurate, the user is prevented from searching again due to the fact that the displayed navigation contents are inaccurate, and the query pressure on a search engine server caused by searching again is reduced; meanwhile, as the more accurate navigation content (the category corresponding to the leaf node) is returned instead of the categories corresponding to all the leaf nodes corresponding to the search word, the number of the search results under the returned navigation content (the category corresponding to the leaf node) is reduced under the condition of ensuring the accuracy, the number of the search results transmitted by the network is reduced, and the return speed and the search efficiency of the search results are improved. By means of the method, the technical problem that in the prior art, the searching efficiency is low due to the fact that the categories corresponding to all the leaf nodes corresponding to the searching words are directly displayed is effectively solved, and the technical effect of improving the searching efficiency is achieved.
It will be apparent to those skilled in the art that the modules or steps of the present application described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.