CN114117253A

CN114117253A - Group search method, device, equipment and storage medium

Info

Publication number: CN114117253A
Application number: CN202111399887.2A
Authority: CN
Inventors: 马晨明; 郝帅; 肖星
Original assignee: Beijing Dajia Internet Information Technology Co Ltd
Current assignee: Beijing Dajia Internet Information Technology Co Ltd
Priority date: 2021-11-19
Filing date: 2021-11-19
Publication date: 2022-03-01

Abstract

The embodiment of the disclosure provides a group search method, a group search device, group search equipment and a storage medium, and belongs to the technical field of the Internet. In the group search method, under the condition that a group search request is initiated by a target account, a plurality of target groups corresponding to the group search request are obtained, then the target groups are sequenced according to the interaction condition of the target account on the target groups, and the sequenced target groups are sent to a terminal where the target account is located. In the sorting process, because the interaction condition of the target account for different groups is considered, the groups which are more in line with the target account searching purpose can be ranked in the front, and the searching result with higher accuracy is obtained to be selected by the target account, so that the experience of application is effectively improved.

Description

Group search method, device, equipment and storage medium

Technical Field

The present disclosure relates to the field of internet technologies, and in particular, to a group search method, apparatus, device, and storage medium.

Background

An Instant Messenger (IM) system can provide services such as real-time communication and data transmission for users, and provides great convenience for the working and life of people. Currently, many IM systems support group chat functionality, where a user wishes to chat in a group, a group search can be performed based on the group name of the group to locate the group.

In the related art, in the process of group search, a search is usually performed in a plurality of groups where a user is located based on a group name input by the user, and the group matching the group name is presented to the user as a search result.

However, as the user group data continuously increases, the same group name is often matched to multiple groups, and in the search result obtained based on the above technical solution, the ranking accuracy of the multiple groups is low, which results in poor user experience.

Disclosure of Invention

The present disclosure provides a group search method, device, electronic device, and storage medium, which can improve the ranking accuracy of a plurality of groups in a search result, and effectively improve the experience of an application. The technical scheme of the disclosure is as follows:

according to a first aspect of the embodiments of the present disclosure, there is provided a group search method, including:

acquiring a plurality of target groups matched with a search keyword based on the search keyword of the group search request of the target account;

acquiring target interaction information of each target group based on the plurality of target groups, wherein the target interaction information indicates the interaction condition of the target account to the target groups;

and sequencing the plurality of target groups based on the value of the target interaction information of each target group, and sending the sequenced plurality of target groups to the terminal where the target account is located.

In the group search method provided by the embodiment of the present disclosure, under the condition that a group search request is initiated by a target account, a plurality of target groups corresponding to the group search request are obtained, then the plurality of target groups are ranked according to the interaction condition of the target account with the plurality of target groups, and the ranked plurality of target groups are sent to a terminal where the target account is located. In the sorting process, because the interaction condition of the target account for different groups is considered, the groups which are more in line with the target account searching purpose can be ranked in the front, and the searching result with higher accuracy is obtained to be selected by the target account, so that the experience of application is effectively improved.

In some embodiments, the obtaining a plurality of target groups based on the search key of the group search request of the target account includes:

based on the search keywords, at least one search mode is applied to search in a plurality of groups where the target account is located, and search results corresponding to each search mode are obtained;

and processing the search result corresponding to each search mode based on the search weight of the target account for each search mode to obtain the plurality of target groups.

By the method, the preference degrees of the user to different search modes are fully considered, the lower limit of the group search quantity can be ensured under the condition that the data volume is continuously increased, and the accuracy of a plurality of target groups is effectively improved.

In some embodiments, the applying at least one search mode based on the search keyword to search in a plurality of groups in which the target account is located to obtain a search result corresponding to each search mode includes:

acquiring a group name corresponding to the search keyword based on the search keyword;

and searching in a plurality of groups where the target account is located based on the group name to obtain the search result.

By the group searching method based on the group name, the searching result matched with the searching keyword can be obtained from the group name, and data support is provided for subsequently determining a plurality of target groups.

In some embodiments, the searching in the plurality of groups where the target account is located based on the group name to obtain the search result includes:

and based on the group name, at least one keyword matching mode is applied to search in a plurality of groups where the target account is located, and the search result is obtained.

By applying the way of multiple keyword matching modes, the range of search results is expanded, more groups which are possibly in line with the search purpose of the target account are obtained, and omission is avoided.

acquiring an account name of a first account corresponding to the search keyword based on the search keyword;

and searching in a plurality of groups in which the target account is located based on the account name of the first account to obtain the search result.

By the mode of searching the group based on the account name, the search result matched with the search keyword can be obtained from the account name, and data support is provided for subsequently determining a plurality of target groups.

In some embodiments, the searching in the plurality of groups in which the target account is located based on the account name of the first account to obtain the search result includes:

acquiring a common group of the target account and the first account based on the account name of the first account;

and sorting the groups in the common group based on the value of the search reference information of each group in the common group to obtain the search result, wherein the search reference information indicates the matching degree between each group in the common group and the group search request and the interaction condition of the target account to the first account.

In this way, the groups in the common group are sorted, so that the groups more suitable for the search purpose of the target account are ranked in the front, and the accuracy of a plurality of target groups obtained subsequently is improved.

In some embodiments, the method further comprises:

obtaining a search log of the target account, wherein the search log indicates historical group search conditions of the target account;

calculating the searching times of the target account for each searching mode based on the searching log;

and obtaining the searching weight of the target account for each searching mode based on the searching times of the target account for each searching mode and the log number of the searching log.

By means of the method for determining the search weight based on the search log, historical group search records of the target account are analyzed, the search weight with high accuracy can be obtained, and therefore the accuracy of a plurality of target groups is improved.

In some embodiments, the method further comprises:

acquiring target interaction information of a first group based on group basic data and group interaction data of the first group, wherein the group basic data indicate basic information of the first group, the group interaction data indicate interaction behaviors of the target account in the first group, and the first group is any one of the plurality of target groups.

The target interaction information of the first group is obtained by combining the group basic data and the group interaction data of the first group, the interaction behavior of the target account is fused with the static information of the first group, the cold start problem of the algorithm is avoided, and the accuracy of the target interaction information is effectively improved.

In some embodiments, the obtaining target interaction information of the first group based on the group basis data and the group interaction data of the first group comprises:

acquiring a plurality of first interaction information of the first group based on the group basic data and the group interaction data of the first group, wherein the first interaction information indicates the interaction condition of the target account to the first group in a corresponding time period;

and obtaining target interaction information of the first group based on the plurality of first interaction information of the first group.

In this way, the interaction condition of the target account to the first group is reflected by different time granularities, and the interaction preference of the target account to the first group in a long term and a short term is fully considered, so that the accuracy of the subsequent calculation of the target interaction information is improved.

In some embodiments, the obtaining target interaction information of the first group based on the plurality of first interaction information of the first group includes:

acquiring a plurality of first weights based on time periods corresponding to the plurality of first interaction information of the first group, wherein the first weights indicate interaction conditions of the target account to any group in the corresponding time period;

and performing weighted summation on the plurality of first interaction information of the first group based on the plurality of first weights to obtain target interaction information of the first group.

The multiple first interaction information is subjected to weighted summation through the multiple first weights, and interaction preference of the target account to the first group under different time granularities is comprehensively considered, so that the accuracy of subsequent target interaction information calculation is improved.

In some embodiments, the method further comprises:

based on the plurality of target groups, acquiring activity information of each target group, wherein the activity information indicates the activity condition of the target account in the target group;

the sorting of the plurality of target groups based on the value size of the target interaction information of each target group, and the sending of the sorted plurality of target groups to the terminal where the target account is located, includes:

and sequencing the plurality of target groups based on the value size of the target interaction information of each target group and the value size of the activity information, and sending the sequenced plurality of target groups to the terminal where the target account is located.

By the mode of combining the target interaction information and the liveness information, the interaction situation and the liveness situation of the target account to different groups are fully considered, and the group expected to be searched by the target account can be ranked in the front, so that a search result with higher accuracy is obtained, and the experience of a user is effectively improved.

In some embodiments, the method further comprises:

and obtaining the activity information of each target group based on the active interaction behavior of the target account in each target group and the occurrence time of the active interaction behavior.

In some embodiments, the method further comprises:

based on the plurality of target groups, acquiring freshness information of each target group, wherein the freshness information indicates the temporary attention degree of the target account to the target group;

and sequencing the plurality of target groups based on the value size of the target interaction information and the value size of the freshness information of each target group, and sending the sequenced plurality of target groups to a terminal where the target account is located.

By the mode of combining the target interaction information and the freshness information, the interaction condition and the temporary attention degree of the target account to different groups are fully considered, and the new group priority can be realized, namely, the group which has interacted with the target account recently is arranged in the front row, so that a search result with higher accuracy is obtained, and the experience of a user is effectively improved.

In some embodiments, the method further comprises:

and obtaining freshness information of each target group based on the target historical interaction behavior of the target account in each target group and the occurrence time of the target historical interaction behavior.

In some embodiments, the method further comprises:

based on the plurality of target groups, acquiring heat information of each target group, wherein the heat information indicates the number of accounts included in the target group;

and sequencing the plurality of target groups based on the value size of the target interaction information of each target group and the value size of the heat information, and sending the sequenced plurality of target groups to the terminal where the target account is located.

By the mode of combining the target interaction information and the popularity information, the interaction condition of the target accounts to different target groups and the number of accounts contained in each target group are fully considered, and small group priority can be realized, so that a search result with high accuracy is obtained, and the experience of a user is effectively improved.

According to a second aspect of the embodiments of the present disclosure, there is provided a group search apparatus, the apparatus including:

the acquisition module is configured to execute a search keyword of a group search request based on a target account, and acquire a plurality of target groups matched with the search keyword;

the acquisition module is configured to acquire target interaction information of each target group based on the plurality of target groups, wherein the target interaction information indicates interaction conditions of the target account with the target groups;

and the sequencing module is configured to execute sequencing on the plurality of target groups based on the value size of the target interaction information of each target group, and send the sequenced plurality of target groups to the terminal where the target account is located.

In some embodiments, the obtaining module comprises:

the searching unit is configured to execute at least one searching mode based on the searching keyword, and search in a plurality of groups where the target account is located to obtain a searching result corresponding to each searching mode;

and the processing unit is configured to execute processing on the search result corresponding to each search mode based on the search weight of the target account for each search mode to obtain the plurality of target groups.

In some embodiments, the search unit is configured to perform:

In some embodiments, the obtaining module is configured to perform:

In some embodiments, the obtaining module comprises:

the first acquisition unit is configured to acquire a plurality of first interaction information of the first group based on the group basic data and the group interaction data of the first group, wherein the first interaction information indicates the interaction condition of the target account to the first group in a corresponding time period;

the second acquisition unit is configured to execute the plurality of first interaction information based on the first group to obtain target interaction information of the first group.

In some embodiments, the second obtaining unit is configured to perform:

In some embodiments, the obtaining module is configured to perform:

the ordering module configured to perform:

In some embodiments, the obtaining module is configured to perform:

the ordering module configured to perform:

In some embodiments, the obtaining module is configured to perform:

the ordering module configured to perform:

According to a third aspect of the embodiments of the present disclosure, there is provided an electronic apparatus including:

one or more processors;

a memory for storing the processor executable program code;

wherein the processor is configured to execute the program code to implement the group search method described above.

According to a fourth aspect of embodiments of the present disclosure, there is provided a computer-readable storage medium including: the program code in the computer readable storage medium, when executed by a processor of an electronic device, enables the electronic device to perform the group search method described above.

According to a fifth aspect of embodiments of the present disclosure, there is provided a computer program product comprising one or more instructions for execution by one or more processors of an electronic device, such that the electronic device is capable of performing the group search method described above.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.

FIG. 1 is a schematic diagram of an implementation environment of a group search method according to an exemplary embodiment;

FIG. 2 is a flow chart of a group search method provided in accordance with an exemplary embodiment;

FIG. 3 is a flow chart of a group search method provided in accordance with an exemplary embodiment;

FIG. 4 is a schematic diagram of a group search method provided in accordance with an exemplary embodiment;

FIG. 5 is a schematic illustration of a collection of log information provided in accordance with an exemplary embodiment;

FIG. 6 is a schematic diagram of a journaling platform provided in accordance with an exemplary embodiment;

FIG. 7 is a diagram illustrating search results for searching a group based on a group name according to an exemplary embodiment;

FIG. 8 is a diagram illustrating a search result from searching a group based on object names in accordance with an illustrative embodiment;

fig. 9 is a schematic structural diagram of a group search apparatus according to an exemplary embodiment;

fig. 10 is a schematic diagram of a server according to an exemplary embodiment.

Detailed Description

In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

The data to which the present disclosure relates may be data that is authorized by a user or sufficiently authorized by parties.

Fig. 1 is a schematic diagram of an implementation environment of a group search method according to an exemplary embodiment, and referring to fig. 1, the implementation environment includes: a terminal 101 and a server 102.

The terminal 101 may be at least one of a smartphone, a smart watch, a desktop computer, a laptop computer, a virtual reality terminal, an augmented reality terminal, a wireless terminal, a laptop computer, and the like, the terminal 101 has a communication function and can access the internet, and the terminal 101 may be generally referred to as one of a plurality of terminals, which is only exemplified by the terminal 101 in the embodiment of the present disclosure. Those skilled in the art will appreciate that the number of terminals described above may be greater or fewer. The terminal 101 may be operated with various applications providing a group search function to the user, such as an IM application, a social contact application, and the like. Illustratively, the terminal 101 is a terminal used by a user, and a user account of the user is registered in an application running in the terminal 101. Taking an example that an IM application program is run on a terminal, the IM application program is used for staff cooperative work inside an enterprise, the IM application program supports a conversation function and a group search function, and when a user wants to chat in a certain group, the user can search and click a group desired by the user through the group search function to realize the conversation function.

The server 102 may be an independent physical server, a server cluster or a distributed file system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like. The server 102 and the terminal 101 may be directly or indirectly connected through wired or wireless communication, which is not limited in the embodiment of the present disclosure. Alternatively, the number of the servers 102 may be more or less, and the embodiment of the disclosure does not limit this. Of course, the server 102 may also include other functional servers to provide more comprehensive and diverse services.

Fig. 2 is a flowchart of a group search method provided according to an exemplary embodiment, and the method is executed by a server as shown in fig. 2, and includes the following steps 201 to 203.

In step 201, the server obtains a plurality of target groups matching a search keyword of a group search request of a target account based on the search keyword.

In the embodiments of the present disclosure, the target account refers to an account that wants to find a desired group through a group search. The search key of the group search request is used to indicate the search purpose of the target account, i.e., the group that the target account desires to search. The target group is matched with the search keyword, and the target group is indicated to be in accordance with the search purpose of the target account. In some embodiments, the search key includes a group name. For example, taking the example that the search key includes a group name "service a", the group names of the target groups are "service a 1", "1 month service a report", and "XX area service a", respectively. In other embodiments, the search key includes an account name of a first account that is in a group with the target account. For example, taking the target account as "user a" and the first account as "user B", the search keyword includes the account name of the first account, i.e., "user a", and the target groups are groups including both "user a" and "user B". In some embodiments, the search key includes a group name and an account name of the first account, which are not limited by the disclosed embodiments.

In step 202, the server obtains target interaction information of each target group based on the plurality of target groups, where the target interaction information indicates interaction conditions of the target account with the target group.

In the embodiment of the present disclosure, each target group has corresponding target interaction information. The target interaction information is used for assisting in predicting the search purpose of the target account, and provides a reference value for subsequently ranking a plurality of target groups, that is, the target interaction information is used for measuring the possibility that the search purpose of the target account is the target group. In some embodiments, the interaction condition of the target account with the target group indicates the interaction preference degree of the target account with the target group, that is, the tendency degree of the target account interacting in the target group. For example, the interaction indicates that the target account sends messages within the target group frequently, and places the target group on top, indicating that the target group is a common group of the target account that prefers to interact within the target group.

In step 203, the server ranks the plurality of target groups based on the value of the target interaction information of each target group, and sends the ranked plurality of target groups to the terminal where the target account is located.

In this embodiment of the present disclosure, the target interaction information is numerical value information, and the server sorts the plurality of target groups according to the value of the target interaction information of each target group, that is, according to the interaction situation of the target account with respect to each target group. For any target group, the larger the target interaction information of the target group is, the higher the interaction condition of the target group indicates the interaction preference degree of the target account for the target group, the higher the possibility that the target group is a search target of the target account is, and the higher the ranking position of the target group is.

The process shown in fig. 2 provides a process in which a server performs an online group search based on a group search request of a target account, and the whole group search process is divided into two stages, i.e., a recall stage and a fine ranking stage, where in the recall stage, i.e., step 201, the server searches a plurality of target groups corresponding to the group search request according to the group search request; in the fine ranking stage, namely step 202 and step 203, the server ranks the groups expected to be searched by the target accounts in the front according to the plurality of target groups and in combination with the interaction situation of the target accounts to different target groups, so that the search result with higher accuracy is obtained. In addition, in some embodiments, the server calls two Remote Procedure Call (RPC) interfaces to respectively implement the two stages of recall and fine ranking, thereby avoiding the mutual influence of recall and fine ranking and improving the accuracy of the search result. It should be noted that fig. 2 is only a basic flow chart of the present disclosure, and the scheme provided by the present disclosure will be further described based on a specific embodiment.

Fig. 3 is a flowchart of a group search method according to an exemplary embodiment, and referring to fig. 3, the interaction subjects of the method are the terminal 101 and the server 102 shown in fig. 1, and the method includes the following steps 301 to 306.

In step 301, the terminal sends a group search request of a target account to the server in response to an operation on the group search page, where the group search request carries a search keyword.

In the embodiment of the present disclosure, a terminal is a terminal used by a user, an account of the user is registered in a target application running on the terminal, and the account is referred to as a target account. Illustratively, a target application program running on a terminal provides a session function and a group search function, the terminal responds to a function selection operation of the target application program and displays a group search page, a user realizes the group search function by implementing operations such as text input and voice input on the group search page, and the terminal responds to the operation on the group search page and sends a group search request of a target account to a server. In some embodiments, a search input box is displayed on the group search page, and a user inputs a search keyword through text input or voice input in the search input box to trigger the terminal to send a group search request of a target account to the server. Illustratively, the target application is an IM application, a social application, or the like, which is not limited in the embodiments of the present disclosure.

In some embodiments, the search key includes a group name. Illustratively, a user corresponding to the target account inputs a group name of a group desired to be searched on the group search page, and the terminal sends a corresponding group search request to the server in response to the input operation. For example, if the user wants to search a group with a group name of "service a", the user inputs a search keyword "service a" on the group search page, thereby triggering the terminal to send a corresponding group search request to the server.

In some embodiments, the search key includes an account name of the first account. Illustratively, a user corresponding to the target account inputs an account name of a first account to be searched on the group search page, and the terminal sends a corresponding group search request to the server in response to the input operation. For example, the target account and the first account "user a" have a plurality of common groups, and then the user corresponding to the target account inputs the search keyword "user a" on the group search page, thereby triggering the terminal to send a corresponding group search request to the server.

In some embodiments, the search key includes a group name and an account name of the first account. Illustratively, a user corresponding to the target account inputs a group name to be searched and an account name of the first account on the group search page, and the terminal sends a corresponding group search request to the server in response to the input operation. For example, the target account and the first account "user a" have a plurality of common groups, and the group name of some common groups in the plurality of common groups is "service a", and as time goes on, the user corresponding to the target account may not accurately recall the name of the common group, but only remember that the target account and the first account "user a" are in the same group, a search keyword "user a" is input on the group search page; service a ", thereby triggering the terminal to send a corresponding group search request to the server.

In some embodiments, the search key comprises a conversation message. Illustratively, a user corresponding to the target account inputs a session message which is desired to be searched on the group search page, and the terminal responds to the input operation and sends a corresponding group search request to the server. For example, in a case that a user corresponding to a target account only remembers specific content of a certain session message, the specific content of the session message can be input on the group search page (e.g., "weather today is really good"), thereby triggering the terminal to send a corresponding group search request to the server.

It should be noted that the above description of the group search request is only an exemplary one, and in some embodiments, the group search request includes any several items of the group name, the account name, and the session message, and the embodiment of the present disclosure does not limit this. In addition, the user corresponding to the target account can trigger the terminal to send a corresponding group search request to the server through any operation mode, which is not limited in the embodiment of the present disclosure.

In step 302, the server receives a group search request for a target account.

In step 303, the server applies multiple search modes based on the search keyword of the group search request of the target account, and searches in multiple groups where the target account is located to obtain a search result corresponding to each search mode.

In the embodiment of the present disclosure, the plurality of search modes include a mode for searching for a group based on a group name, a mode for searching for a group based on an account name, a mode for searching for a group based on a session message, and the like, and the embodiment of the present disclosure does not limit the kind of the search mode.

In some embodiments, the server calls an RPC interface, recalls from a search engine (e.g., an elastic search, ES) the group names of the groups in which the target account is located based on the account identification (e.g., user id) of the target account; and based on the search keywords, applying a plurality of search modes to search in a plurality of groups where the target account is located to obtain search results corresponding to each search mode.

In some embodiments, after receiving the group search request, the server performs search preprocessing on the group search request, and searches in a plurality of groups in which the target account is located by applying a plurality of search modes according to a preprocessing result to obtain a search result corresponding to each search mode. For example, the search preprocessing includes word segmentation, length interception, and intention analysis of search keywords of the group search request. The intention analysis is to determine the preference degree of the target account for different search modes based on the group search habit of the target account, so that different weights can be set according to different search modes in a targeted manner when search results corresponding to each search mode are fused to obtain a plurality of target groups in the follow-up process, and therefore the plurality of target groups with higher accuracy are obtained.

In some embodiments, the server synchronously applies multiple search modes based on the group search request, and searches in multiple groups where the target account is located to obtain a search result corresponding to each search mode. By the synchronous execution mode, the search Response Time (RT) can be reduced as much as possible, and the search efficiency is improved. In some embodiments, the server sequentially applies multiple search modes based on the group search request, and searches in multiple groups where the target account is located to obtain a search result corresponding to each search mode. By the sequential execution mode, the load pressure of the server can be reduced as much as possible, and the service life of the server is prolonged.

The various search modes described above are described below based on several examples.

First, a mode of searching for a group based on a group name.

Illustratively, the process of the server obtaining the corresponding search result based on the mode includes the following two steps:

step 1, the server obtains the group name corresponding to the search keyword based on the search keyword.

And the server extracts the group name in the search keyword based on the search keyword of the group search request. For example, the group name is "search". It should be noted that, in this mode, the server may perform semantic analysis on the search keyword to extract the group name, or may directly use the search keyword as the group name, and the like, which is not limited in this embodiment of the disclosure.

And 2, searching in a plurality of groups of the target account by the server based on the group name to obtain the search result.

The server obtains a plurality of groups where the target account is located based on the account identifier (such as user id) of the target account, and obtains the search result based on the matching degree between the group name and the group names of the plurality of groups. Illustratively, the server calculates similarity between the group name and the group names of the plurality of groups, and takes the group with the similarity greater than or equal to a first threshold as the search result, for example, the first threshold is a preset threshold of 50%, which is not limited in the embodiments of the present disclosure.

In some embodiments, the server searches a plurality of groups in which the target account is located by applying at least one keyword matching pattern based on the group name to obtain the search result. The keyword matching mode comprises Chinese matching, Chinese pinyin full matching, prefix matching, partial matching, Chinese simple pinyin full matching and the like. Referring to table 1, in table 1, taking a group name corresponding to a search keyword as "search" as an example, showing several keyword forms of the group name, and the server obtains a search result corresponding to a pattern of searching the group based on the group name based on a matching degree between the several keywords of the group name shown in table 1 and the group names of a plurality of groups in which the target account is located. By applying the way of matching the multiple keywords with the pattern, the requirement of multi-pattern search can be supported, and the accuracy of the search result can be improved, for example, by continuously referring to table 1, under the condition that the user corresponding to the target account inputs ss, the corresponding group can be positioned to be searched according to the ss.

TABLE 1

Chinese character	Complete spelling	Simple spelling	Full spelling and simple spelling	Prefix	In part
						Searching	sousuo	ss	sous	sou	suo

In some embodiments, the number of groups in the search result is less than or equal to a first number threshold, different keyword matching patterns correspond to different matching weights, the server searches a plurality of groups in which the target account is located based on the group name, the multiple keyword matching patterns, and the matching weight corresponding to each keyword matching pattern to obtain the search result, where the matching weight indicates a preference degree of the target account for the keyword matching pattern, and the first number threshold may be set according to an actual requirement, for example, the first number threshold is 1000, which is not limited in the embodiments of the present disclosure. In some embodiments, the server stores matching pattern information of each account, where the matching pattern information includes a matching weight of each account for each keyword matching pattern, and the server obtains the matching pattern information of the target account by using an account identifier (e.g., user id) of the target account as an index, and obtains the matching weight of the target account for each keyword matching pattern. In some embodiments, the server processes the log information of the target account every first time interval (for example, the first time interval is 1 day) to obtain matching pattern information of the target account, and stores the matching pattern information, so that when a group search request is received, the matching pattern information of the target account is obtained according to the group search request to obtain a matching weight of the target account for each keyword matching pattern (this process is the same as the process of obtaining the search weight of the target account for each search pattern by the server according to the search log of the target account in subsequent step 304, and therefore details are not repeated here). Of course, in some embodiments, the matching weight can be set in other manners, for example, the server determines the preference degree of the target account for each keyword matching pattern by performing intent analysis on the group search request, and obtains the corresponding matching weight. For example, the matching weight is adjusted in time according to actual needs, and the embodiment of the present disclosure does not limit the manner in which the server obtains the matching weight.

Illustratively, taking the multiple keyword matching patterns including chinese matching, partial matching, and prefix matching as an example, the matching weights of the target account for each keyword matching pattern are 60%, 30%, and 10%, respectively, and if the result obtained based on the multiple keyword matching patterns includes 1500 groups, the number of groups obtained through the chinese matching pattern is 500, the number of groups obtained through the partial matching pattern is 600, and the number of groups obtained through the prefix matching pattern is 400. In some examples, the number of groups in the search results is less than or equal to the first number threshold, and the results are processed based on the product of the first number threshold and each matching weight, wherein the server processes the groups obtained by the chinese matching mode, taking 500 groups (since 1000 × 60% ═ 600>500, all groups obtained by chinese matching are desirably added to the search results). The server processes the groups obtained by the partial matching mode, and takes the group ordered at the top 300 bits (since 1000 × 30% ═ 300<600, the group ordered at the top 300 bits of the 600 groups obtained by partial matching is added to the search result). The server processes the groups obtained through the prefix matching mode, the groups ranked at the top 100 bits are taken (because 1000 × 10% is 100<400, 400 groups obtained through prefix matching are added to the search result, and the finally obtained search result comprises 900 groups, that is, by the method, a large number of groups obtained by applying various keyword matching modes are controlled to be below a first quantity threshold (screened from the quantity 1500 to 900, which is smaller than the first quantity threshold 1000), so that the number of the groups in the search result is effectively controlled according to the matching weight of the target account to each keyword matching mode, the subsequent data processing amount is reduced, and the group search efficiency is improved.

In other embodiments, the number of groups in the search result is equal to the first number threshold, and if the number of groups in the result obtained according to one keyword matching pattern satisfies the first number condition, the later-ranked groups in the groups corresponding to the keyword matching patterns including more groups are supplemented to the results with the insufficient number. The first quantity condition is that the product of the matching weight corresponding to the keyword matching pattern and the first quantity threshold is greater than the group quantity in the result corresponding to the keyword matching pattern.

For example, continuing to take the multiple keyword matching patterns including chinese matching, partial matching, and prefix matching as an example, the matching weights of the target account for each keyword matching pattern are 60%, 30%, and 10%, respectively, if the result obtained based on the multiple keyword matching patterns includes 1500 groups, the number of groups in the search result is equal to the first number threshold 1000, where the number of groups obtained by the chinese matching pattern is 500, the number of groups obtained by the partial matching pattern is 600, the number of groups obtained by the prefix matching pattern is 400, and the results are processed based on the product of the first number threshold and each matching weight, where the server processes the groups obtained by the chinese matching pattern, takes all 500 groups (since 1000 × 60% ═ 600>500, meets the requirement, adds all 500 groups obtained by chinese matching to the search result), it can be seen that the number of groups in the result obtained according to the chinese matching mode satisfies the number condition (i.e. 600 groups obtained through the chinese matching mode should be searched, but only 500 groups, but no 100 groups, may also be understood as 100 empty positions in the result obtained based on the chinese matching mode to be supplemented, or more groups may be included in the search result corresponding to other matching modes). In this case, the server processes the groups obtained by the partial matching mode, taking the groups ordered at 350 bits (since 1000 × 30% ═ 300<600, the 600 groups obtained by partial matching, the group ordered at the top 350 bits are added to the search results, and the 50 groups ordered at 301 to 350 bits can be understood as being supplemented to the results obtained by the chinese matching mode). The server processes the groups obtained by the prefix matching mode, and takes the group ordered at the top 150 bits (since 1000 × 10% ═ 100<400, 400 groups obtained by prefix matching are added to the search result, and the 50 groups arranged at the top 150 bits can be understood as being supplemented to the result obtained by the chinese matching mode), and the final search result includes 1000 groups. It should be noted that, for example, the illustration is only an exemplary illustration, in some embodiments, the number of groups corresponding to the finally obtained chinese matching, partial matching, and prefix matching patterns may also be 500, 380, and 120, respectively, and the like, which is not limited in this disclosure. By the method, a large number of groups obtained by applying various keyword matching modes are controlled to be within the first quantity threshold (screened from the quantity 1500 to 1000, which is equal to the first quantity threshold 1000), so that the quantity of the groups in the search results is effectively controlled according to the matching weight of the target account to each keyword matching mode, the subsequent data processing amount is reduced, and the group search efficiency is improved.

It should be understood that the examples are merely illustrative, and the server may also obtain the search result corresponding to each search pattern in other manners, and in some embodiments, in the case that only one group is obtained according to one keyword matching pattern, the search result of the keyword matching pattern does not need to be processed based on the matching weight, which is not limited by the embodiments of the present disclosure.

By the searching mode based on at least one keyword matching mode and corresponding matching weight, the lower limit of group searching quality can be guaranteed and the group which is more in line with the searching habit of the target account can be selected according to the preference degree of the target account to different keyword matching modes under the condition that the data volume is continuously increased, so that the accuracy of a plurality of target groups is effectively improved, meanwhile, the number of the groups in the searching result is controlled, the subsequent data processing amount is reduced, and the group searching efficiency is improved.

Second, a mode of searching groups based on account names.

step 1, the server obtains the account name of the first account corresponding to the search keyword based on the search keyword.

And the server extracts the account name of the first account in the search keyword based on the search keyword of the group search request. For example, the account name of the first account is "zhang san". It should be noted that, in this mode, the server may perform semantic analysis on the search keyword to extract the account name of the first account, may also directly use the search keyword as the account name of the first account, and the like, which is not limited in this embodiment of the disclosure. In addition, the examples herein are merely illustrative, and the embodiments of the present disclosure do not limit the types of the search keywords.

And 2, searching in a plurality of groups where the target account is located based on the account name of the first account to obtain the search result.

The server obtains a plurality of groups where the target account is located based on the account identifier (such as user id) of the target account, obtains the account identifier (such as user id) of the first account based on the account name of the first account, obtains a plurality of groups where the first account is located based on the account identifier of the first account, and obtains the search result based on the common group of the target account and the first account. In some embodiments, after obtaining the plurality of groups in which the target account is located based on the account identifier of the target account, the server takes, as a search result, a group including both the target account and the first account in the plurality of groups based on account information in the plurality of groups. That is, the server may obtain an intersection after obtaining multiple groups where the two accounts are located, or may search for a group that includes the first account in the multiple groups after obtaining multiple groups where the target account is located, which is not limited in this embodiment of the present disclosure.

In some embodiments, the server obtains a plurality of pieces of key information of the first account based on the account name of the first account, and searches a plurality of groups in which the target account is located based on the plurality of pieces of key information to obtain the search result. Wherein the plurality of key information is used to identify the account name of the first account from different perspectives. For example, referring to table 2, taking the account name of the first account as "zhang san" as an example, the plurality of key information includes a full spelling, a simple spelling, a full spelling + a simple spelling, a prefix, a part, a remark, a nickname, a user name, and the like, which is not limited in this embodiment of the disclosure.

In some embodiments, the process of the server searching in the plurality of groups in which the target account is located based on the plurality of key information includes: and the server searches in a plurality of groups where the target account is based by applying a plurality of keyword matching modes or any one of the keyword matching modes based on the plurality of key information to obtain the search result. That is, taking table 2 as an example, the account name of the first account is "zhang san", the plurality of pieces of key information include full pinyin, simple pinyin, full pinyin + simple pinyin, prefix, part, remark, nickname, user name, and the like, and the plurality of keyword matching patterns include: full pinyin match, simple pinyin match, full pinyin + simple pinyin match, prefix match, partial match, remark match, nickname match, and username match, among others. The server obtains a search result corresponding to the pattern of searching the group based on the account name based on the matching degree between the account names included in the plurality of groups where the target account is located based on several kinds of key information of the account name shown in table 2. By applying the way of multiple keyword matching modes, the requirement of multi-mode search can be supported, and the accuracy of the search result can be improved, for example, with reference to table 2, under the condition that zs is input into the target account, the group of the first account "zhang san" can be located according to zs.

In addition, in the mode of searching the group based on the account name, different keyword matching patterns correspond to different matching weights according to a mode similar to the mode of searching the group based on the group name, that is, the server searches in a plurality of groups where the target account is located based on the account name of the first account and based on at least one keyword matching pattern and the matching weight corresponding to each keyword matching pattern, and obtains the search result. The specific process may refer to the above mode of searching for the group based on the group name, and the embodiments of the present disclosure are not described herein again.

TABLE 2

In some embodiments, the server obtains a common group of the target account and the first account based on the account name of the first account; and sorting the groups in the common group based on the value of the search reference information of each group in the common group to obtain the search result, wherein the search reference information indicates the matching degree between each group in the common group and the group search request and the interaction condition of the target account to the first account. The matching degree between each group and the group search request indicates the possibility that a first account exists in the corresponding group, and the interaction condition of the target account with the first account indicates the interaction preference degree of the target account with the first account, namely the interaction tendency degree of the target account with the first account. For example, the interaction indicates that the target account often sends a message, a click session, a search session, etc. to the first account, which is not limited by the embodiment of the present disclosure. Illustratively, the search reference information is numerical value information, and is calculated by the following formula (1):

Group_Score(u，v，k)＝α₁HitKeyword(k)+α₂FavUser(u，v) (1)

in the formula, Group _ Score represents search reference information of a Group (i.e., Group Score), u represents a target account, v represents a first account, and k represents a search keyword (i.e., of the first account)Account name, or any one of the information of the first account), HitKeyword indicates the degree of matching between the group and the group search request, FavUser indicates the interaction of the target account with the first account, and α₁And alpha₂Respectively represent preset weights, and it should be noted that α₁And alpha₂The method can be adjusted according to actual requirements, and the embodiment of the disclosure does not limit the method. In this way, the groups in the common group are sorted, so that the groups more suitable for the search purpose of the target account are ranked in the front, and the accuracy of a plurality of target groups obtained subsequently is improved.

Through the mode of searching the group based on the account name, the functional requirement that the target account carries out group search through the account name can be supported, and the experience of application is effectively improved. For example, as time goes by, a user corresponding to a target account may not be able to accurately recall the name of a group, but only remember that the user corresponding to a certain account (e.g., XXX) is in the same group, and in the case that the user corresponding to the target account inputs a search keyword "XXX", the server is able to locate the group that the user corresponding to the target account wants to search for.

And a third mode of searching for a group based on a session message.

step 1, the server obtains the conversation information corresponding to the search keyword based on the search keyword.

And the server extracts the session message in the search keyword based on the search keyword of the group search request. For example, the conversation message is "weather today is really good". It should be noted that, in this mode, the server may perform semantic analysis on the search keyword to extract the session message, may also directly use the search keyword as the session message, and the like, which is not limited in this embodiment of the disclosure. In addition, the examples herein are merely illustrative, and the embodiments of the present disclosure do not limit the types of the search keywords in the group search request.

And 2, searching in the historical conversation messages of a plurality of groups in which the target account is positioned based on the conversation messages to obtain the search result.

The server obtains a plurality of groups where the target account is located based on an account identifier (such as a user id) of the target account, and obtains the search result based on a matching degree between the session message and historical session messages of the plurality of groups. Illustratively, the server acquires the historical conversation message of each group, and takes the group corresponding to the historical conversation message containing the conversation message as the search result. Of course, the server may obtain the search result in other ways, for example, the server obtains the historical session message of each group within 7 days, so as to obtain the corresponding search result, and the like, which is not limited in this disclosure.

In addition, in the mode of searching the group based on the session message, the server can apply at least one keyword matching mode based on the session message according to a mode similar to the mode of searching the group based on the group name to search the historical session messages of a plurality of groups in which the target account is located, and obtain the search result. For example, taking the conversation message as "today's weather is really good", the plurality of keyword matching patterns include patterns of chinese matching (today's weather is really good), chinese pinyin full matching (jintianianqizhenhao), prefix matching (today), partial matching (weather is really good), chinese simple pinyin full matching (jttqzh), and the like. Similarly, in some embodiments, different keyword matching patterns correspond to different matching weights, that is, the server searches historical conversation messages of a plurality of groups in which the target account is located based on the conversation message and based on at least one keyword matching pattern and the matching weight corresponding to each keyword matching pattern, and obtains the search result. The specific process may refer to the above mode of searching for the group based on the group name, and the embodiments of the present disclosure are not described herein again.

It should be noted that the above several search modes are merely illustrative, and in some embodiments, the server can apply other search modes to obtain corresponding search results, for example, a mode of searching for a group based on group identification, and the like, which is not limited by the embodiments of the present disclosure. In addition, in the several search modes, the server can search by applying various keyword matching modes and corresponding matching weights, so that the lower limit of group search quality can be ensured and groups more conforming to the search habit of the target account can be selected according to the preference degree of the target account to different keyword matching modes under the condition that the data volume is continuously increased, the accuracy of a plurality of target groups is effectively improved, meanwhile, the number of the groups in the search result is controlled by the mode, the subsequent data processing amount is reduced, and the group search efficiency is improved.

In step 304, the server processes the search result corresponding to each search mode based on the search weight of the target account for each search mode, so as to obtain a plurality of target groups.

In the embodiment of the present disclosure, the search result corresponding to each search mode includes at least one group, and the server processes the groups in the search results according to the search weight of the target account for each search mode to obtain the plurality of target groups. Wherein the search weight of the target account for each search mode indicates a degree of preference of the target account for different search modes.

In some embodiments, the number of the plurality of target groups is less than or equal to the second number threshold, and the server processes the search result corresponding to each search mode based on the search weight of the target account for each search mode, so as to obtain a plurality of target groups. The second quantity threshold may be set according to an actual requirement, for example, the second quantity threshold is 200, which is not limited in this disclosure. Illustratively, taking 100 groups as an example (the number of groups for which the server obtains the corresponding group based on each search mode may be set according to actual needs, which is not limited by the embodiments of the present disclosure), the second number threshold is 200, the plurality of search modes include a group name search group based mode, an account name search group based mode, and a session message search group based mode, and the search weights of the target account for each search mode are 60%, 20%, and 20%, respectively. The server processes the search results corresponding to the mode of searching the group based on the group name, and takes all 100 groups (because 200 × 60% ═ 120>100, the group corresponding to the mode of searching the group based on the group name is all regarded as the target group). The server processes the search results corresponding to the pattern for searching the group based on the account name, and takes the group with the top 40 bits (since 200 × 20% ═ 40<100, the group with the top 40 bits of the 100 search results obtained by the pattern for searching the group based on the account name is taken as the target group). The server processes the search results corresponding to the pattern of the session message search group, and takes the group with the top 40 bits (since 200 × 20% ═ 40<100, the group with the top 40 bits of the 100 search results obtained by the pattern of the session message search group is taken as the target group), and finally 180 target groups are obtained. That is, by this method, the number of groups obtained by applying multiple search modes is controlled to be below the second number threshold (screened from number 300 to 180, which is smaller than the second number threshold 200), so that the number of multiple target groups is effectively controlled according to the search weight of each search mode by the target account, and the accuracy of the multiple target groups is improved.

In other embodiments, the number of the target groups is equal to a second number threshold, and when the number of groups in the result obtained according to one search mode satisfies a second number condition, the next-ranked groups in the groups corresponding to the search modes including more groups are supplemented to the result with the insufficient number. The second quantity condition is that the product of the search weight corresponding to the search pattern and the second quantity threshold is greater than the number of groups in the result corresponding to the search pattern. It should be noted that this process is the same as the process of obtaining corresponding search results by applying multiple keyword matching patterns and matching weights, so that it is not specifically described here, but only by following an exemplary example, for example, if the above example is continued, where each search result includes 100 groups, the second quantity threshold is 200, the search weights of the target account for each search pattern are 60%, 20%, and 20%, respectively, then the server processes the search results corresponding to the pattern based on the group name search group, and takes all 100 groups (there are 20 slots to be supplemented); processing the search results corresponding to the mode of searching the group based on the account name, and taking the group ranked at the top 50 bits (supplementing the group ranked at 41-50 bits to the search results corresponding to the mode of searching the group based on the group name); and processing the search results corresponding to the mode of searching the group based on the session message, and taking the group with the top 50 bits (the same way), thereby finally obtaining 200 target groups. That is, by this method, the groups obtained by applying the multiple search modes are controlled to be the second quantity threshold (screened from the quantity 300 to 200, which is equal to the second quantity threshold 200), so that the quantity of the multiple target groups is effectively controlled according to the search weight of the target account for each search mode, and the accuracy of the multiple target groups is improved.

It should be understood that the examples are merely illustrative, and the server may also use other ways to obtain the plurality of target groups. In some embodiments, in a case that a search result only includes one group, the search result is not processed, and the group is directly used as a target group, which is not limited in the embodiments of the present disclosure.

In some embodiments, the server stores the search information of each account, the search information including the search weight of each account for each search mode, and in this step 304, the server obtains the search weight of each account for each search mode from the search information of each account based on the account identification (e.g., user id) of the target account. In some embodiments, after the server performs search preprocessing on the group search request in step 303, a preprocessing result is obtained, where the preprocessing result includes a search weight of the target account for each search mode, that is, the server determines a preference degree of the target account for each search mode by performing intent analysis on the group search request, so as to obtain a corresponding search weight, and the embodiment of the present disclosure does not limit a manner in which the server obtains the search weight.

In some embodiments, the server obtains the search weight of the target account for each search mode according to the search log of the target account. Illustratively, this process includes the following steps:

step 1, the server obtains a search log of the target account, wherein the search log indicates the historical group search condition of the target account.

And the historical group search record of the target account is stored in the server in the form of a search log. For example, the terminal where the target account is located generates a corresponding search log based on any group search request of the target account, and submits the corresponding search log to the server.

And 2, the server calculates the search times of the target account for each search mode based on the search log.

The server determines the search mode adopted by the target account in the historical group search process based on the search log of the target account, and calculates the search times of the target account for each search mode. In some embodiments, the server updates the search logs of the target account every day, counts the number of searches of the target account for each search mode in the target duration, for example, the server updates the search logs of the target account at zero point every day, and counts the number of searches of the target account for each search mode in the past 90 days, and in some embodiments, this process is also understood as the server performs T +1 statistics on the search logs of the target account, which is not limited by the embodiments of the present disclosure.

And 3, the server obtains the search weight of the target account for each search mode based on the search times of the target account for each search mode and the log number of the search logs.

Taking any search mode as an example, the server obtains the search weight of the target account for the search mode based on the ratio between the number of times of search of the target account for the search mode and the number of logs of the search log. For example, if the log number of the search log of the target account is 10, the number of searches of the target account for the mode of searching the group based on the group name is 8, and the number of searches of the target account for the mode of searching the group based on the account name is 2, the search weights of the target account for the two search modes are 80% and 20%, respectively, indicating that the target account tends to search the group by the group name.

By means of the method for determining the search weight based on the search log, the historical group search records of the target account are analyzed, the search weight with high accuracy can be obtained, the accuracy of a plurality of target groups can be effectively improved under the condition that with the continuous increase of business and under the large data scale, and particularly, the accuracy of the plurality of target groups can be effectively improved under the condition that the number of recall groups is large due to the fact that the content indicated by the search keywords input by the target account is small (for example, the search keywords are 'ss').

It should be noted that, the process of generating the search weight may be performed in advance by the server, that is, the search weight is obtained by the server and then stored in the search information of the target account, so that when a new group search request is received, the corresponding search weight is obtained according to the account identifier (such as the user id) of the target account; the process of generating the search weight may also be performed when the server performs the search preprocessing on the group search request when performing the step 303, which is not limited in the embodiment of the present disclosure.

In addition, in the embodiment of the disclosure, after obtaining the search result corresponding to each search mode, the server processes the search result corresponding to each search mode based on the search weight of the target account for each search mode to obtain a plurality of target groups. In other embodiments, after obtaining the search result corresponding to each search mode, the server directly integrates the search results to obtain the plurality of target groups, which is not limited in this disclosure.

Through the above steps 303 and 304, the server applies a plurality of search modes based on the group search request of the target account, and obtains a plurality of target groups corresponding to the group search request. By the method, the preference degree of the user to different search modes is fully considered, the lower limit of the group search quality can be ensured under the condition that the data volume is continuously increased, the group which is more in line with the search habit of the target account is selected, and therefore the accuracy of a plurality of target groups is effectively improved. It should be noted that, in some embodiments, the server may also apply multiple keyword matching patterns and corresponding matching weights to perform a search in the process of applying multiple search patterns, and in this way, the preference degrees of the user for different search patterns and the preference degrees of the user for different keyword matching patterns are fully considered, so that under the condition that the data volume is continuously increased, the lower limit of the group search quality is further ensured, a group more conforming to the search habits and the search input habits of the target account is selected, and the accuracy of multiple target groups is greatly improved.

In some embodiments, the server applies any one of a plurality of search modes based on the group search request of the target account, searches among a plurality of groups in which the target account is located to obtain a search result corresponding to the search mode, and processes the search result corresponding to the search mode based on the search weight of the target account for the search mode to obtain a plurality of target groups corresponding to the group search request. Illustratively, in this case, the search weight of the target account for the search pattern may be 1, which is not limited by the embodiment of the present disclosure.

In some embodiments, the server applies any of multiple search modes based on the group search request of the target account, searches among multiple groups where the target account is located, and obtains a search result corresponding to each search mode, thereby obtaining multiple target groups, and it should be noted that this process is the same as step 303 and step 304 described above, and therefore is not described herein again.

In step 305, the server obtains target interaction information of each target group based on the plurality of target groups, wherein the target interaction information indicates interaction conditions of the target account with the target group.

In the embodiment of the disclosure, the server obtains the target interaction information of each target group based on the group identifier (e.g. group id) of each target group and the account identifier (e.g. user id) of the target account. In some embodiments, the server stores target interaction information of a plurality of groups in which the target account is located, and the server obtains the target interaction information of each target group from the target interaction information of the plurality of groups in which the target account is located, using the group identifier of each target group and the account identifier of the target account as indexes. In some embodiments, the server processes the log information of the target account every second time interval to obtain and store target interaction information of a plurality of groups where the target account is located, so that when a group search request is received, the target interaction information of each target group is obtained according to the group search request. For example, the second period of time is 1 day. By the method, the interaction preference degree of the target account for each target group can be determined according to the target interaction information of each target group, so that which target groups belong to the group of which the target account keeps long-term stable communication is judged, and the accuracy of sequencing of a plurality of subsequent target groups is improved.

In the following, taking any one of the plurality of target groups as an example, a process of processing the log information of the target account in advance by the server to obtain the target interaction information of the group is introduced. It should be understood that the target group belongs to a plurality of groups in which the target account is located, and thus, the following process is applicable to any one of the plurality of groups in which the target account is located. In addition, in the embodiment of the present disclosure, the example is described in which the server generates and stores the target interaction information of the group in advance, and in other embodiments, the server may further calculate the target interaction information of each target group in real time after acquiring a plurality of target groups, which is not limited in the embodiment of the present disclosure.

Illustratively, any one of the plurality of target groups is referred to as a first group, and the server obtains target interaction information of the first group based on group basis data and group interaction data of the first group, wherein the group basis data indicates basis information of the first group, and the group interaction data indicates interaction behavior of the target account in the first group. Wherein, the basic information of the first group comprises the account number, the group creation time, the group label and the like in the first group; the interactive behavior of the target account in the first group includes sending a message, collecting, presenting, clicking on the group, staying time in the group, setting/canceling the setting, opening/canceling the do-not-disturb, moving in/out of a message box, dismissing the group, transferring the group, searching for a message, adding a robot, exiting the group, and the like. It should be noted that, the embodiment of the present disclosure does not limit the basic information of the first group and the interaction behavior of the target account in the first group.

In some embodiments, the process of acquiring, by the server, the target interaction information of the first group based on the group basis data and the group interaction data of the first group includes the following two steps:

step 1, the server acquires a plurality of first interaction information of the first group based on the group basic data and the group interaction data of the first group, wherein the first interaction information indicates the interaction condition of the target account to the first group in a corresponding time period.

And each first interaction message has a corresponding time period. That is, the plurality of first interaction information reflect the interaction of the target account with the first group through different time granularities. For example, the first interaction information corresponds to the time periods of 7 days and 30 days, respectively. It should be understood that the 7 days and 30 days are only illustrative, and in some embodiments, the setting can be performed according to actual requirements, for example, 60 days and 10 days, and for example, 7 days, 30 days and 60 days, and the like, and the embodiment of the disclosure does not limit the present disclosure.

It should be noted that the group basic data and the group interaction data are obtained by processing the log information of the target account. For example, referring to table 3, the server performs Statistics based on the log information of the target account, the time granularity is divided into 30 days and 7 days, which are respectively calculated once per day as long-term and short-term interaction preferences of the target account, and StatTime (where Stat is an abbreviation of staticiscs Statistics) in the table indicates the time of 0 o' clock on the day. The log information of the target account is counted through different time granularities, and the long-term and short-term interaction preference of the target account can be fully considered, so that the accuracy of the subsequent calculation of the target interaction information is improved.

TABLE 3

In some embodiments, the group basis data includes first basis information and second basis information, wherein the first basis information indicates a number of designated accounts in a first group, e.g., the designated accounts are leaders, for example, the first group is a work group. The second basic information indicates a creation time of the first group. It should be noted that, in some embodiments, the group basic data further includes other basic information, which is not limited by the embodiments of the present disclosure.

In some embodiments, the group interaction data includes first interaction reference information, second interaction reference information, third interaction reference information, fourth interaction reference information, and fifth interaction reference information. Wherein the first interactive reference information indicates how far the occurrence time of the message sent by the target account in the first group is from the current time (e.g., as embodied by a weighted message sending score (weightedSendsmsg)); the second interaction reference information indicates how far the target account clicks the first group from the current time (e.g., as embodied by weighted group click score); third interactive reference information indicates an average dwell time of the target account within the first group; the fourth interaction reference information indicates the preference degree of the target account for a plurality of target interaction behaviors of the first group, wherein the plurality of target interaction behaviors comprise message sending, set-top/set-top canceling, opening/set-free disturbance canceling, message box moving in/out, group dismissal, group transfer, message search, group quitting, robot adding and the like; and the fifth interaction reference information indicates the occurrence time of the last interaction behavior of the target account to the first group.

For example, the group interaction data of the first group refers to table 4, where table 4 shows the statistics of the target account in the last N days of the first group, where N is a positive integer. The "last operation time" refers to an occurrence time of a last interaction behavior of the target account with respect to the first group, for example, if the last interaction behavior of the target account with respect to the first group is turn-on do-not-disturb, the "last operation time" indicates that the last interaction behavior is turn-on do-not-disturb, the occurrence time is XX minutes (represented by a timestamp in 13-bit millisecond units) in XX month and XX month in XXXX year, and so on, which is not limited by the embodiment of the disclosure.

TABLE 4

Illustratively, the first interactive reference information corresponds to "message sending score with weight" in table 4; the second interactive reference information corresponds to the "click score with weight group" in table 4; the third interactive reference information corresponds to "average stay time in group" in table 4; the fourth interactive reference information corresponds to the "set-top/set-top canceling times, open/set-disturbance canceling times, message box moving in/out times, dismissal group, transfer group, message search times, exit group, and add robot" in table 4; the fifth interactive reference information corresponds to "last operation time" in table 4. It should be noted that the fourth interactive reference information is also matched with the content items corresponding to Positive and negative in the table, where Positive represents the Positive interactive behavior of the target account with respect to the first group, and negative represents the negative interactive behavior of the target account with respect to the first group, and by this way, the unified calculation is performed on the partial interactive behaviors, and the calculation resources are saved. Of course, in some embodiments, the adjustment may also be performed according to actual requirements, for example, any several kinds of interaction behaviors indicated by the fourth interaction reference information are calculated in a unified manner, and for example, a certain interaction behavior indicated by the fourth interaction reference information is calculated separately, which is not limited in this disclosure.

The first interactive reference information and the second interactive reference information are both numerical value information, and the first interactive reference score and the second interactive reference score are calculated through the following formula (2), so that the first interactive reference information and the second interactive reference information are obtained:

wherein t represents the current time, a_tAnd indicating the occurrence time of the interactive behavior of the target account (such as the occurrence time of a message or the occurrence time of a click group), wherein both delta and bias are hyper-parameters (which can be adjusted according to actual requirements), and accumulating the scores of each time of the interactive behavior of the target account to obtain corresponding interactive reference information. It should be noted that formula (2) above is a way to calculate the score based on the calculation rule of time decay, and taking the first interactive reference information as an example, the closer the occurrence time of the message sent by the target account is to the current time, the larger the first interactive reference information is.

It should be noted that, the embodiment of the present disclosure takes formula (2) as an example for description, and in some embodiments, the first interactive reference information and the second interactive reference information can be determined by other calculation methods according to the same principle, which is not limited by the embodiment of the present disclosure.

Illustratively, taking the various interaction reference information shown in table 4 above as an example, the server obtains the first interaction information of the first group by the following formula (3):

where u represents the target account, g represents the first group, S (u, g) represents the first interaction information, WeightedSendMsg represents first interaction reference information, WeightedClick represents second interaction reference information, RemainTime represents third interaction reference information, and sum (positive) -sum (negative) represents fourth interaction reference information, namely, positive interaction behavior and negative interaction behavior are respectively accumulated and summed, last_optimeDenotes fifth interactive reference information, LeaderNum denotes first basic information (i.e., the number of leaders) in the group basic data, GroupCreateTime denotes second basic information (i.e., group creation time) in the group basic data, ω₁To omega₇The weight corresponding to each kind of information is respectively represented. In addition, in the formula (3), various information needs to be standardized so that the reference value is controlled to be in the same order of magnitude, and at the same time, the scaling is performed by the log function, so that the excessive increase can be avoided, and the stability of the algorithm can be ensured. In some embodiments, in the weight corresponding to each of the above information, a part of the weight may be set to be 0, that is, a part of the information is selected to participate in calculating the first interaction information, which is not limited in the embodiment of the present disclosure.

Through the method, the group interaction data of the first group are classified, namely, the interaction behaviors corresponding to the first interaction reference information, the second interaction reference information, the third interaction reference information and the fifth reference information are used as main information, the interaction behavior corresponding to the fourth interaction reference information is used as secondary information, and meanwhile, the group basic data of the first group are also classified, so that the weight corresponding to each information is reasonably set under the condition of cold start of an algorithm or insufficient sample data. In addition, the log information of the target account and the static information of the first group are fused in the manner of calculating the first interactive information shown in the formula (3), so that the cold start problem of the algorithm is further avoided, and the accuracy of the first interactive information is effectively improved.

And 2, the server obtains target interaction information of the first group based on the plurality of first interaction information of the first group.

The server acquires a plurality of first weights based on time periods corresponding to a plurality of first interaction information of a first group, wherein the first weights indicate interaction conditions of the target account to any group in the corresponding time period; and carrying out weighted summation on a plurality of first interaction information of the first group based on the plurality of first weights to obtain target interaction information of the first group.

Illustratively, in the case that the plurality of first interaction information respectively correspond to the time periods of 7 days and 30 days, the target interaction score of the first group is calculated by the following formula (4), so as to obtain the target interaction information of the first group:

w(u，g)＝β₁S₃₀(u，g)+β₂S₇(u，g) (4)

in equation (4), u represents the target account, g represents the first group, S₃₀(u, g) first interaction information corresponding to 30 days, S₇(u, g) represents first interaction information corresponding to 7 days, β₁And beta₂First weights corresponding to 30 days and 7 days are indicated, respectively. It should be noted that the plurality of first weights can be adjusted according to actual needs, for example, data of the user in the last 30 days is more stable than data of the user in the last 7 days, and data of the last 7 days is more capable of reflecting recent interest and preference of the user, β is used₁And beta₂Are set to 30% and 70%, respectively, which are not limited by the embodiments of the present disclosure. It should be understood that the above formula (4) is only an exemplary one, and in the case that the amount of the first interaction information is larger, the above formula (4) is adapted to obtain the corresponding target interaction information.

Through the step 305, the server acquires the target interaction information of each target group based on the plurality of target groups, provides a reference value for subsequently sorting the plurality of target groups, and can assist in predicting the search purpose of the target account.

In step 306, the server ranks the plurality of target groups based on the value of the target interaction information of each target group, and sends the ranked plurality of target groups to the terminal where the target account is located.

In the embodiment of the present disclosure, the server ranks the target groups with large target interaction information in front based on the value size of the target interaction information of each target group, and obtains a plurality of sorted target groups. In some embodiments, the server obtains, based on the plurality of target groups, activity information, freshness information, and heat information of each target group, or obtains any one of the foregoing information, or obtains any several of the foregoing information, and combines the target interaction information to sort the plurality of target groups, which is not limited in this disclosure. It should be noted that this alternative embodiment will be described in detail in the following embodiments, and will not be described herein again.

After receiving the sorted target groups, the terminal where the target account is located arranges and displays the target groups according to the sorting of the target groups, so that the target group which best meets the target account searching purpose is preferentially displayed to the target account.

In some embodiments, the server sends the sorted target groups to the terminal where the target account is located after performing target processing on the sorted target groups. Illustratively, the target processing includes aggregation ordering, classification labels, and concatenation group names, among others. Wherein, the aggregating and sorting means that the target groups hit multiple search keywords in the multiple target groups are ranked in front, for example, the group search request includes "XX area service a; for example, if the user a, the group name of the target group 1 is "XX area service a", the target group 1 includes the user a, and the group name of the target group 2 is "XX area service a", but the target group 2 does not include the user a, the server ranks the target group 1 in front of the target group 2. The category marking is to mark the hit search keyword so that the terminal where the target account is located can highlight the hit search keyword according to the marking, for example, marking the name of a red group, or marking the remark of a red user. The splicing group name is obtained by splicing the target group without the group name according to the account name of each account in the target group, so that the terminal where the target account is located displays the group name after being relied. For example, taking an example that a certain target group includes "user a", "user B", and "user C", the group names obtained by server concatenation are "user a, user B, and user C". By means of the method for processing the targets of the plurality of sequenced target groups, the display layer of the search results is optimized, the form of displaying the plurality of target groups by the terminal where the target account is located is enriched, and user experience is effectively improved.

The following describes an alternative embodiment of the server in step 306 combining the target interaction information, the activity information, the freshness information, and the heat information of each target group to order a plurality of target groups according to several situations.

First, the server sorts the plurality of target groups based on target interaction information and liveness information of each target group.

Wherein the activity information indicates activity of the target account in the target group. For example, the active situation is embodied by active interaction behaviors of the target account such as sending messages, forms, collections, and set-top in the target group.

Illustratively, the process of the server sorting the plurality of target groups includes the following two steps:

step 1, the server obtains the activity information of each target group based on the plurality of target groups.

Wherein, the server obtains the activity information of each target group based on the group identification (such as group id) of each target group and the account identification (such as user id) of the target account. In some embodiments, the server stores activity information of a plurality of groups in which the target account is located, and the server obtains the activity information of each target group by using the group identifier of each target group and the account identifier of the target account as indexes. In some embodiments, the server processes the log information of the target account every third duration to obtain and store activity information of a plurality of groups in which the target account is located, so that when a group search request is received, the activity information of the plurality of target groups is obtained according to the group search request. For example, the third time period is 1 h. By the method, the activity degree of each target group by the target account can be determined according to the activity information of each target group, so that which target groups belong to the communication groups which are active in the short term of the target account can be judged, and the accuracy of sequencing of a plurality of subsequent target groups can be improved.

In some embodiments, the server obtains the activity information of each target group based on the active interaction behavior of the target account in each target group and the occurrence time of the active interaction behavior. Illustratively, the server calculates the liveness score of each target group based on the above formula (2), so as to obtain the liveness information of each target group, which is not described herein again.

In addition, it should be noted that, in the present embodiment, a timing for the server to obtain the activity information is not limited, in some embodiments, the server performs the step of obtaining the activity information before performing the step 305, in other embodiments, the server performs the step of obtaining the activity information simultaneously while performing the step 305, and in still other embodiments, the server performs the step of obtaining the activity information after performing the step 305.

And 2, the server sorts the plurality of target groups based on the value size of the target interaction information and the value size of the activity information of each target group, and sends the sorted plurality of target groups to the terminal where the target account is located.

The server performs weighted summation on the target interaction information and the liveness information of each target group based on a first sorting weight corresponding to the target interaction information and a second sorting weight corresponding to the liveness information to obtain first sorting reference information of each target group, sorts the plurality of target groups based on the first sorting reference information, and sends the sorted plurality of target groups to a terminal where a target account is located. For any target group, the larger the target interaction information and the activity information of the target group are, the larger the first ordering reference information of the target group is, and the more advanced the ordering of the target group is. In some embodiments, the first sorting weight and the second sorting weight are preset weights and can be adjusted according to actual requirements, for example, both the first sorting weight and the second sorting weight are 50%, which is not limited in the embodiments of the present disclosure.

And secondly, the server sorts the target groups based on the target interaction information and freshness information of each target group.

Wherein the freshness information indicates a temporary level of interest to the target group by the target account. In some embodiments, the freshness information indicates the target account's level of interest in the target group in a short period of time, e.g., the level of interest in the target group in the last 1 hour. In addition, in some embodiments, the temporary attention degree is represented by the occurrence time of the last interactive behavior of the target account in the target group, or the temporary attention degree is represented by historical interactive behaviors such as group creation time of the target group, set-top of the target group by the target account, no disturbance, conference group participation, time of the target account entering the target group, and search click, and such historical interactive behaviors can indicate whether the target account needs to pay attention to the target group. Of course, in some embodiments, the temporary attention level can also be represented by other historical interaction behaviors, which is not limited by the embodiments of the present disclosure.

step 1, the server acquires freshness information of each target group based on the plurality of target groups.

Wherein the server obtains freshness information of each target group based on a group identification (e.g., group id) of each target group and an account identification (e.g., user id) of the target account. In some embodiments, the server obtains freshness information of each target group based on the target historical interaction behavior of the target account in each target group and the occurrence time of the target historical interaction behavior. In some embodiments, the target historical interaction behavior includes active interaction behavior and passive interaction behavior, and the active interaction behavior includes, for example, in a target group: after a target account inputs a certain search keyword, the target group is clicked in a search result, the target account actively enters the target group, the target group is a conference group, and the like. The passive interaction behavior comprises: target accounts are invited into the target group and target accounts are mentioned within the target group (e.g., by @), and so on. Illustratively, the server obtains the occurrence time of the target historical interaction behavior of the target account in each target group based on the group identifier of each target group and the account identifier of the target account, for example, the freshness degree information is numerical value type information, and calculates the freshness degree score of each target group through the following formula (5), thereby obtaining the freshness degree information of each target group.

In equation (5), FreshScore represents the freshness score, u represents the target account, n_tIndicates the current time, a_tRepresenting the time of occurrence, theta, of the historical interaction behavior of the object₁And theta₂Respectively represent the weight, and can be input according to the actual requirementThe line adjustment, 3600000, is in milliseconds and represents 1h, that is, in the case that the time of occurrence of the target historical interaction behavior is less than 1h from the current time, the freshness score θ is taken₁And under the condition that the distance between the occurrence time of the target historical interaction behavior and the current time is less than 3h, the freshness score is taken as theta₂And under the condition that the distance between the occurrence time of the target historical interaction behavior and the current time is more than 3h, the freshness score is 0. It should be noted that the formula (5) is only shown schematically, and in some embodiments, the freshness score of the target group can be adjusted according to actual needs, which is not limited in the embodiments of the present disclosure.

In addition, it should be noted that the timing for the server to acquire the freshness information is not limited in the embodiments of the present application, in some embodiments, the server performs the step of acquiring the freshness information before performing step 305, in other embodiments, the server performs the step of acquiring the freshness information simultaneously while performing step 305, and in still other embodiments, the server performs the step of acquiring the freshness information after performing step 305.

And 2, the server sorts the plurality of target groups based on the value size of the target interaction information and the value size of the freshness information of each target group, and sends the sorted plurality of target groups to the terminal where the target account is located.

The server performs weighted summation on the target interaction information and the freshness information of each target group based on a third sorting weight corresponding to the target interaction information and a fourth sorting weight corresponding to the freshness information to obtain second sorting reference information of each target group, sorts the plurality of target groups based on the second sorting reference information, and sends the sorted plurality of target groups to a terminal where a target account is located. For any target group, the larger the target interaction information and freshness information of the target group are, the larger the second sorting reference information of the target group is, and the further the sorting of the target group is. In some embodiments, the third sorting weight and the fourth sorting weight are preset weights and can be adjusted according to actual requirements, for example, the third sorting weight and the fourth sorting weight are 60% and 40%, respectively, which is not limited in the embodiments of the present disclosure.

By the mode of combining the target interaction information and the freshness information, the interaction condition and the temporary attention degree of the target account to different groups are fully considered, and the priority of a new group can be realized, namely, the target account which has interacted recently and the group which needs attention are arranged in the front, so that a search result with higher accuracy is obtained, the real-time performance of the search is reflected, and the experience of the application is effectively improved.

And thirdly, the server sorts the target groups based on the target interaction information and the heat information of each target group.

The popularity information indicates the number of accounts included in the target group, and the popularity information is smaller when the number of accounts included in the target group is larger. In some embodiments, the heat information indicates a number of messages within the target group, the greater the number of messages within the target group, the smaller the heat information. In some embodiments, the popularity information indicates the number of accounts included in the target group and the number of messages in the target group, that is, in the case that the number of accounts included in the target group is large (e.g., greater than 1000) and the number of messages in the target group is large (e.g., greater than 10000), the target group is a large group or a water group, and is likely not to be a search purpose of the target account.

step 1, the server acquires heat information of each target group based on the plurality of target groups.

Wherein, the server obtains the hot degree information of each target group based on the group identification (such as group id) of each target group. In some embodiments, the server stores the popularity information of a plurality of groups in which the target account is located, and the server obtains the popularity information of each target group by using the group identifier of each target group as an index. In some embodiments, the server counts and stores the heat information of each group every fourth time interval, so that when a group search request is received, the heat information of a plurality of target groups is obtained according to the group search request. For example, the fourth period of time is 1 day. By the method, which target groups belong to the large group or the water group can be determined according to the heat information of each target group, so that the sequencing accuracy of a plurality of subsequent target groups is improved.

In addition, it should be noted that, in the embodiment of the present application, a timing for the server to acquire the heat information is not limited, in some embodiments, the server performs the step of acquiring the heat information before performing the step 305, in other embodiments, the server performs the step of acquiring the heat information simultaneously while performing the step 305, and in still other embodiments, the server performs the step of acquiring the heat information after performing the step 305.

And 2, the server sorts the plurality of target groups based on the value size of the target interaction information and the value size of the heat information of each target group, and sends the sorted plurality of target groups to the terminal where the target account is located.

The server performs weighted summation on the target interaction information and the heat information of each target group based on a fifth sorting weight corresponding to the target interaction information and a sixth sorting weight corresponding to the heat information to obtain third sorting reference information of each target group, sorts the plurality of target groups based on the third sorting reference information, and sends the sorted plurality of target groups to a terminal where a target account is located. For any target group, the larger the target interaction information and the heat information of the target group are, the larger the third sorting reference information of the target group is, and the more the sorting of the target group is. In some embodiments, the fifth and sixth sorting weights are preset weights and can be adjusted according to actual requirements, for example, the fifth and sixth sorting weights are 80% and 20%, respectively, which is not limited in the embodiments of the present disclosure.

By the mode of combining the target interaction information and the popularity information, the interaction condition of the target account to different target groups and the account number or the message number (or the combination of the two) contained in each target group are fully considered, and the small group priority can be realized, so that a search result with higher accuracy is obtained, and the experience of a user is effectively improved.

In addition, in the above process, the heat information of each target group may be understood as global heat information of the target group, that is, the heat information is used to indicate the heat of the target group itself, or the interest degree of any account in the target group. In some embodiments, the server ranks the plurality of target groups in combination with personalized heat information, global heat information, and target interaction information of each target group, where the personalized heat information indicates a degree of interest of a target account in the target group, and is embodied by an interaction behavior of the target account in the target group, for example, the interaction behavior includes sending a message, searching for a click, collecting, and the like, which is not limited by the embodiments of the present disclosure. Illustratively, this process includes the following two steps:

step 1, the server acquires target heat information of each target group based on the plurality of target groups.

The target heat information is obtained by processing the global heat information and the personalized heat information of each target group. Illustratively, the server obtains global heat information and personalized heat information of each target group based on a group identifier (e.g., group id) of each target group and an account identifier (e.g., user id) of a target account, and processes the global heat information and the personalized heat information of each target group to obtain target heat information of each target group. For example, taking the global heat information and the personalized heat information as numerical class information, the server takes the sum of the global heat information and the personalized heat information of the target group as the target heat information of the target group. That is, in the case that the target group is a big group or a water group, if the target account has interacted with the target group recently, it is indicated that the target account is likely to be interested in the target group currently, and the global heat information of the target group may be increased, so that the big group or the water group interested by the target account can be prevented from being arranged to the end in the subsequent sorting, and the accuracy of the sorting result is improved. Of course, in some embodiments, the server may also process the global heat information and the personalized heat information in other manners, which is not limited in this disclosure.

In some embodiments, the server stores global heat information and personalized heat information of a plurality of groups in which the target account is located, and the server obtains the global heat information and the personalized heat information of each target group by using the group identifier of each target group and the account identifier of the target account as indexes. In some embodiments, the server counts and stores the global heat information of each group every fifth time interval, for example, the fifth time interval is 1 day. The server counts and stores the personalized heat information of each group every sixth time interval, for example, the sixth time interval is 1 hour, which is not limited in the embodiment of the present disclosure. By the method, the interest degree of the target accounts in the target groups can be comprehensively considered on the basis of determining which target groups belong to a large group or a water group according to the target heat information of each target group, so that the sequencing accuracy of a plurality of subsequent target groups is improved.

And 2, the server sorts the target groups based on the value size of the target interaction information and the value size of the target heat information of each target group, and sends the sorted target groups to the terminal where the target account is located.

The process of sequencing the target groups is the same as the process of sequencing the target groups by the server based on the target interaction information and the heat information of each target group, and therefore the process is not repeated herein.

By the method of combining the target interaction information and the target heat information, the interaction condition of the target account to different target groups, the account number or the message number (or the combination of the two) contained in each target group and the current interest degree of the target account to the target group are fully considered, and the large group or the water group interested by the target account can be prevented from being arranged to the end on the basis of realizing the priority of the small group, so that a search result with higher accuracy is obtained, and the experience of application is effectively improved.

Fourthly, the server sorts the target groups based on the target interaction information, the activity information, the freshness information and the heat information of each target group.

The server obtains the activity information, the freshness information, and the heat information of each target group based on the group identifier (e.g., group id) of each target group and the account identifier (e.g., user id) of the target account (this process is the same as the above process, and is not described here again). The server performs weighted summation on the target interaction information, the activity information, the freshness information and the heat information of each target group based on a seventh sorting weight corresponding to the target interaction information, an eighth sorting weight corresponding to the activity information, a ninth sorting weight corresponding to the freshness information and a tenth sorting weight corresponding to the heat information to obtain fourth sorting reference information of each target group, sorts the target groups based on the fourth sorting reference information, and sends the sorted target groups to a terminal where the target account is located. For any target group, the larger the target interaction information, the activity information, the freshness information and the heat information of the target group are, the larger the fourth sorting reference information of the target group is, and the further the sorting of the target group is. In some embodiments, the seventh sorting weight, the eighth sorting weight, the ninth sorting weight, and the tenth sorting weight are all preset weights, and can be adjusted according to actual requirements, for example, the seventh sorting weight, the eighth sorting weight, the ninth sorting weight, and the tenth sorting weight are respectively 50%, 20%, and 10%, which is not limited in the embodiments of the present disclosure.

By the mode of combining the target interaction information, the activity information, the freshness information and the heat information, the interaction condition, the activity condition and the temporary attention degree of the target accounts to different target groups and the account number or the message number contained in each target group (or the combination of the two) are fully considered, multiple factors can be comprehensively considered, and the most expected groups searched by the target accounts are arranged in the front, so that the search result with high accuracy is obtained, and the experience of a user is greatly improved.

It should be noted that, in some embodiments, a policy that the server sorts the plurality of target groups may be adjusted according to actual needs, for example, the plurality of target groups may be sorted by combining any several items of the target interaction information, the activity information, the freshness information, and the heat information, and a specific implementation process is the same as the above process, and therefore is not described herein again, and the embodiment of the present disclosure does not limit how the server specifically combines the target interaction information, the activity information, the freshness information, and the heat information.

The group search method provided by the embodiment of the present disclosure is schematically described below based on the embodiments shown in fig. 2 and fig. 3, and with reference to fig. 4 to fig. 9 described below.

Fig. 4 is a schematic diagram of a group search method according to an exemplary embodiment. As shown in fig. 4, the group search method includes the following four stages: the method comprises the steps of log information collection, log information processing, target interaction information generation and online group search, wherein the four steps are connected in sequence to form a service closed loop.

The log information collection is to collect corresponding log information based on the interaction behavior of the target account to each group by implementing log burial in the target application program, so as to provide data support for the subsequent processes of calculating interaction information and the like. For example, referring to fig. 5, fig. 5 is a schematic diagram of a log information collection provided according to an example embodiment. As shown in fig. 5, taking the target application as an IM application as an example, the interaction behavior of the target account in the group is logged, which mainly includes entering the group, leaving the group, sending messages in the group, staying time, table state, setting top, avoiding disturbance, moving in/out of a message box, adding collection, and the like, and the interaction behaviors feed back the interaction preference degree of the target account to the current group from the side. Of course, the illustration in the drawings is only schematic, and a developer can adjust the journal burying point according to actual needs, which is not limited by the embodiment of the disclosure. In some embodiments, taking the IM application program as an example for an internal enterprise employee collaboration office, after implementing log embedding in the target application program, data is uniformly collected into a Hive (a Hadoop-based data warehouse tool) traffic base table of an enterprise, and each business department extracts respective log information for a business representation and finally falls into a database table of the department. The log information, Hive, query and the like are managed uniformly through the log embedded platform, and data service is provided in a one-stop mode. Schematically, the journaling platform is shown in fig. 6, and fig. 6 is a schematic diagram of a journaling platform according to an exemplary embodiment.

The log information processing and target interaction information generation means that collected log information is processed, and interaction behavior indexes of the target accounts in the group under different time dimensions are calculated in an aggregation mode. This process can refer to step 305, table 3 and table 4 in the embodiment shown in fig. 3, which are not described herein again. In some embodiments, the server periodically calculates the target interaction information of the target account once a day, thereby determining which target groups belong to the communication group with the target account stable for a long time, and improving the accuracy of the sequencing of the plurality of subsequent target groups.

The online group search refers to a process that, when a group search request is initiated by a target account, a server acquires a plurality of target groups of the target account based on the group search request, sorts the target groups, and finally returns a search result to a terminal where the target account is located. This process is referred to from step 301 to step 306, and therefore will not be described herein.

After receiving the sorted target groups, the terminal where the target account is located can display the target groups in an arrangement manner according to the sequence of the target groups. Generally, whether the group search index is accurate enough is measured by the click rate of the online search result, taking the example that the IM application is used for the collaborative office of the employees in the enterprise, after the IM application is optimized based on the group search method provided by the embodiment of the disclosure, the click rate of Top3 of the online group search has reached 92.01%, and the click rate of Top10 is closer to 98.12%. Schematically, as shown in fig. 7, fig. 7 is a schematic diagram of a search result for searching a group based on a group name, after a service a is searched by the group name, the search result ranks all members of the common group service a, a service a-a group, a service a & calendar service, etc. in the first few digits. As shown in fig. 8, fig. 8 is a schematic diagram of a search result based on an account name search group according to an exemplary embodiment, and when the group is searched by an account name, a communication group that is most recently related is also ranked in the front.

In summary, when a group search request is initiated by a target account, a plurality of target groups corresponding to the group search request are obtained, then the plurality of target groups are ranked according to the interaction situation of the target account with the plurality of target groups, and the ranked plurality of target groups are sent to a terminal where the target account is located. In the sorting process, because the interaction condition of the target account for different groups is considered, the groups which are more in line with the target account searching purpose can be ranked in the front, and the searching result with higher accuracy is obtained to be selected by the target account, so that the experience of application is effectively improved.

Fig. 9 is a schematic structural diagram of a group search apparatus according to an exemplary embodiment. As shown in fig. 9, the apparatus includes an obtaining module 901 and an ordering module 902.

An obtaining module 901 configured to execute a search keyword of a group search request based on a target account, and obtain a plurality of target groups matched with the search keyword;

the obtaining module 901 is configured to perform obtaining target interaction information of each target group based on the plurality of target groups, where the target interaction information indicates interaction conditions of the target account with the target group;

the sorting module 902 is configured to execute sorting the plurality of target groups based on the value size of the target interaction information of each target group, and send the sorted plurality of target groups to the terminal where the target account is located.

In some embodiments, the obtaining module 901 includes:

In some embodiments, the search unit is configured to perform:

In some embodiments, the obtaining module 901 is configured to perform:

In some embodiments, the obtaining module 901 includes:

In some embodiments, the second obtaining unit is configured to perform:

In some embodiments, the obtaining module 901 is configured to perform:

the ordering module 902 is configured to perform:

In some embodiments, the obtaining module 901 is configured to perform:

the ordering module 902 is configured to perform:

In some embodiments, the obtaining module 901 is configured to perform:

the ordering module 902 is configured to perform:

It should be noted that: in the group search apparatus provided in the above embodiment, only the division of each functional module is illustrated when performing the group search, and in practical applications, the function distribution may be completed by different functional modules according to needs, that is, the internal structure of the apparatus is divided into different functional modules to complete all or part of the functions described above. In addition, the group search apparatus and the group search method provided in the above embodiments belong to the same concept, and specific implementation processes thereof are described in detail in the method embodiments and are not described herein again.

An embodiment of the present disclosure further provides an electronic device, including:

one or more processors;

a memory for storing the processor executable program code;

wherein the processor is configured to execute the program code to implement the processes executed by the server in the group search method provided by the above-mentioned embodiments of the method.

In some embodiments, the program code related to the embodiments of the present application may be deployed to be executed on one electronic device or on multiple electronic devices located at one site, and the multiple electronic devices distributed at multiple sites and interconnected by a wired network or a wireless network may form a block chain system.

Taking an electronic device as an example of a server, fig. 10 is a schematic structural diagram of a server according to an exemplary embodiment, where the server 1000 may generate a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 1001 and one or more memories 1002, where the one or more memories 902 store at least one program code, and the at least one program code is loaded and executed by the one or more processors 1001 to implement the processes executed by the server in the group search method provided by the above-mentioned method embodiments. Of course, the server 1000 may also have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input and output, and the server 1000 may also include other components for implementing the functions of the device, which are not described herein again.

In an exemplary embodiment, a computer readable storage medium comprising program code, such as a memory 1002 comprising program code, executable by a processor 1001 of the server 1000 to perform the group search method described above is also provided. Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a Random Access Memory (RAM), a compact-disc read-only memory (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.

In an exemplary embodiment, a computer program product is also provided that includes one or more instructions for execution by one or more processors of an electronic device to enable the electronic device to perform the group search method described above.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A group search method, comprising:

acquiring a plurality of target groups matched with a search keyword based on the search keyword of a group search request of a target account;

acquiring target interaction information of each target group based on the plurality of target groups, wherein the target interaction information indicates the interaction condition of the target object to the target groups;

and sequencing the plurality of target groups based on the value size of the target interaction information of each target group, and sending the sequenced plurality of target groups to a terminal where the target account is located.

2. The group search method according to claim 1, wherein the obtaining a plurality of target groups matching the search keyword based on the search keyword of the target account group search request comprises:

3. The group search method according to claim 2, wherein the applying at least one search mode based on the search keyword to search a plurality of groups in which the target account is located to obtain a search result corresponding to each search mode comprises:

4. The group search method according to claim 3, wherein the searching among the plurality of groups in which the target account is located based on the group name to obtain the search result comprises:

and based on the group name, applying at least one keyword matching mode to search in a plurality of groups where the target account is located to obtain the search result.

5. The group search method according to claim 2, wherein the applying at least one search mode based on the search keyword to search a plurality of groups in which the target account is located to obtain a search result corresponding to each search mode comprises:

and searching in a plurality of groups where the target account is located based on the account name of the first account to obtain the search result.

6. The group search method according to claim 5, wherein the searching among the plurality of groups in which the target account is located based on the account name of the first account to obtain the search result comprises:

obtaining a common group of the target account and the first account based on the account name of the first account;

and sorting the groups in the common groups based on the value of the search reference information of each group in the common groups to obtain the search results, wherein the search reference information indicates the matching degree between each group in the common groups and the group search request and the interaction condition of the target account to the first account.

7. A group search apparatus, comprising:

8. An electronic device, characterized in that the electronic device comprises:

one or more processors;

a memory for storing the processor executable program code;

wherein the processor is configured to execute the program code to implement the group search method of any one of claims 1 to 6.

9. A computer-readable storage medium, wherein program code in the computer-readable storage medium, when executed by a processor of an electronic device, enables the electronic device to perform the group search method of any of claims 1 to 6.

10. A computer program product comprising one or more instructions for execution by one or more processors of an electronic device to enable the electronic device to perform the group search method of any of claims 1-6.