Search

CN-117171707-B - User intention information identification method, device, computer equipment and storage medium

CN117171707BCN 117171707 BCN117171707 BCN 117171707BCN-117171707-B

Abstract

The invention relates to the technical field of computers and discloses a method, a device, computer equipment and a storage medium for identifying user intention information, wherein the method for identifying the user intention information comprises the steps of determining a weight value according to click behavior information of a target user clicking a click object on a target application; the method comprises the steps of determining a first intention value according to a weight value, matching search word information of a target user with a preset first term, determining a plurality of bias labels of the target user according to a first matching result, determining a second intention value according to search times corresponding to the bias labels, matching chat record information of the target user with a preset second term, determining a plurality of intention labels of the target user according to a second matching result, determining a third intention value according to the occurrence times corresponding to the intention labels, and generating user intention information according to the first intention value, the second intention value and the third intention value. The method and the device generate the user intention more accurately.

Inventors

  • XU JIAN

Assignees

  • 北京自如信息科技有限公司

Dates

Publication Date
20260508
Application Date
20230920

Claims (9)

  1. 1. A method for identifying user intent information, the method comprising: determining the weight values of a plurality of click objects according to click behavior information of the click objects of target users on a click target application, and determining the first intention values corresponding to a plurality of attributes according to the weight values of the plurality of click objects, wherein the first intention values corresponding to different attributes are the same or different; The search term information of the target user is matched with a first entry which is preset, a plurality of bias labels of the target user are determined according to a first matching result, and a second intention value is determined according to the search times corresponding to the bias labels; matching the chat record information of the target user with a second entry which is preset, determining a plurality of intention labels of the target user according to a second matching result, and determining a third intention value according to the occurrence times corresponding to the plurality of intention labels; generating user intention information according to the first intention value, the second intention value and the third intention value; Before the user intention information is generated according to the first intention value, the second intention value and the third intention value, matching the plurality of attributes, the plurality of bias labels and the plurality of intention labels to obtain a plurality of intention information, wherein one intention information corresponds to at least three intention values; the method comprises the steps of generating user intention information according to the first intention value, the second intention value and the third intention value, carrying out weighted summation on at least three intention values in each intention information to obtain a plurality of matching values, wherein one intention information corresponds to one matching value, screening out the matching value which is larger than or equal to a first preset value from the plurality of matching values, and generating user intention information according to the intention information corresponding to the matching value.
  2. 2. The method of claim 1, wherein prior to determining the weight values for the plurality of click objects based on click behavior information of the target user clicking on a click object on a target application, the method further comprises: And filtering click behavior information of a target user from a user behavior log, wherein the user behavior log comprises records of target application operation of the target user, and the click behavior information comprises click information of various click objects.
  3. 3. The method according to claim 2, wherein determining the weight values of the plurality of click objects according to the click behavior information of the click objects on the target application by the target user comprises: determining the clicking times corresponding to the multiple clicking objects respectively according to the clicking behavior information of the target user; And determining weight values of the multiple clicking objects according to the clicking times respectively corresponding to the multiple clicking objects, wherein the weight values are positively correlated with the clicking times.
  4. 4. The method according to claim 2, wherein determining the first intent value corresponding to the plurality of attributes according to the weight values of the plurality of click objects comprises: sorting the plurality of click objects according to the weight values to obtain a first sorting queue; sequentially selecting a first preset number of multiple click objects from the head of the first sequencing queue; matching corresponding attributes for a plurality of click objects, wherein one click object corresponds to one attribute; And determining the first intention values of the various attributes according to the weight values of the various click objects in a first preset quantity in the first sorting queue.
  5. 5. The method according to claim 1 or 2, wherein determining the second intent value according to the search times corresponding to the plurality of bias labels comprises: Sorting the plurality of deflection labels according to the search times corresponding to the plurality of deflection labels to obtain a second sorting queue; Sequentially selecting a second preset number of the plurality of deflection labels from the head of the second sequencing queue; And determining the second intention values of the plurality of biased tags according to the searching times of the plurality of biased tags with a second preset number in the second sequencing queue.
  6. 6. The method according to claim 1 or 2, wherein determining a third intent value according to the number of occurrences corresponding to the plurality of intent tags comprises: sequencing the plurality of intention labels according to the occurrence times corresponding to the plurality of intention labels to obtain a third sequencing queue; Sequentially selecting a third preset number of the plurality of intention tags from the head of the queue in the third sequencing queue; And determining the third intention values of the plurality of intention tags according to the occurrence times of the plurality of intention tags with a third preset number in the third sorting queue.
  7. 7. An apparatus for identifying user intention information, the apparatus comprising: the weight value determining module is used for determining the weight values of a plurality of click objects according to the click behavior information of the click objects of the target user on the click target application; The first intention value determining module is used for determining first intention values corresponding to various attributes according to the weight values of the click objects, wherein the first intention values corresponding to different attributes are the same or different; the first matching module is used for matching the search word information of the target user with a first entry which is preset; the bias label determining module is used for determining a plurality of bias labels of the target user according to the first matching result; The second intention value determining module is used for determining a second intention value according to the search times corresponding to the plurality of deflection labels; the second matching module is used for matching the chat record information of the target user with a second entry which is preset; The intention label determining module is used for determining a plurality of intention labels of the target user according to the second matching result; The third intention value determining module is used for determining a third intention value according to the occurrence times corresponding to the plurality of intention labels; The user intention information determining module is used for generating user intention information according to the first intention value, the second intention value and the third intention value; Before the user intention information is generated according to the first intention value, the second intention value and the third intention value, matching the plurality of attributes, the plurality of bias labels and the plurality of intention labels to obtain a plurality of intention information, wherein one intention information corresponds to at least three intention values; the method comprises the steps of generating user intention information according to the first intention value, the second intention value and the third intention value, carrying out weighted summation on at least three intention values in each intention information to obtain a plurality of matching values, wherein one intention information corresponds to one matching value, screening out the matching value which is larger than or equal to a first preset value from the plurality of matching values, and generating user intention information according to the intention information corresponding to the matching value.
  8. 8. A computer device, comprising: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform the method of identifying user intention information as claimed in any one of claims 1 to 6.
  9. 9. A computer-readable storage medium having stored thereon computer instructions for causing a computer to execute the method of identifying user intention information according to any one of claims 1 to 6.

Description

User intention information identification method, device, computer equipment and storage medium Technical Field The present invention relates to the field of computer technologies, and in particular, to a method and apparatus for identifying user intention information, a computer device, and a storage medium. Background At present, population mobility is increased, house renting and selling requirements are increased, and in the house renting and selling process, a traditional house recommending method is that through communication with a customer, the intention of the customer is determined, and a corresponding house source is searched according to the intention of the customer. However, this method is time-consuming and laborious, and requires detailed conversations with clients, and in the course of conversations, the obtained client intention information is not accurate enough, and in the related art, when a house source is found for a client, the house source is recommended to the client according to the search record of the client, but when the scope of client search is wide, the intention information of the client is not accurately mastered, and the recommended house source does not meet the intention of the client. Disclosure of Invention In view of the above, the present invention provides a method, apparatus, computer device and storage medium for identifying user intention information, so as to solve the problem of inaccurate grasp of the intention information of a client. The invention provides a user intention information identification method, which comprises the steps of determining weight values of multiple clicking objects according to clicking behavior information of the clicking objects on a target application of a target user, determining first intention values corresponding to the multiple clicking objects according to the weight values of the multiple clicking objects, determining first intention values corresponding to different attributes, matching search word information of the target user with preset first terms, determining multiple bias tags of the target user according to a first matching result, determining second intention values according to search times corresponding to the multiple bias tags, matching chat record information of the target user with preset second terms, determining multiple intention tags of the target user according to a second matching result, determining third intention values according to the occurrence times corresponding to the multiple intention tags, and generating user intention information according to the first intention values, the second intention values and the third intention values. The method and the device have the advantages that in the embodiment of the invention, the weight values of various click objects are determined according to the click behavior information of the click objects on the target application of the target user, the first intention values corresponding to various attributes are determined according to the weight values of various click objects, and the first intention values corresponding to different attributes are the same or different. The method comprises the steps of matching search word information of a target user with a preset first term, determining a plurality of deflection labels of the target user according to a first matching result, and determining a second intention value according to search times corresponding to the deflection labels. Matching chat record information of the target user with a second entry which is preset, determining a plurality of intention labels of the target user according to a second matching result, and determining a third intention value according to the occurrence times corresponding to the plurality of intention labels. According to the first intention value, the second intention value and the third intention value, user intention information is generated, the three intention values are fused, and the user intention information is generated. In an alternative embodiment, before determining the weight values of the multiple click objects according to the click behavior information of the click objects on the target application by the target user, the method further comprises filtering the click behavior information of the target user from a user behavior log, wherein the user behavior log comprises records of the operation of the target application by the target user, and the click behavior information comprises click information of the multiple click objects. In an alternative implementation mode, the weight values of multiple types of click objects are determined according to the click behavior information of the target user on the click objects on the target application, wherein the weight values of the multiple types of click objects are determined according to the click behavior information of the target user, the click times corresponding to the multiple types of click obje