Method for determining answer weight, answer determining method, device and storage medium

A technology for determining methods and answers, applied in the computer field, can solve the problem that the weight of the answer cannot accurately reflect the true satisfaction of the user's answer, and achieve the effect of improving the user experience.

Active Publication Date: 2021-08-06
MIGU CO LTD +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of this, the main purpose of the present invention is to provide an answer weight determination method, answer determination method, device and storage medium, which solves the problem that the answer weight cannot accurately reflect the user's true satisfaction with the answer

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for determining answer weight, answer determining method, device and storage medium
  • Method for determining answer weight, answer determining method, device and storage medium
  • Method for determining answer weight, answer determining method, device and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0115] Embodiment 1: The user consults the machine customer service to obtain the target answer

[0116] I. User U1 contacts the machine customer service S1, talks to the machine customer service, and asks the machine customer service "Which express is it?";

[0117] II. The robot customer service searched two candidate answers according to the user's questions, which are recorded as A1 and A2 respectively. Initially, the weights of A1 and A2 are both 0;

[0118] III. The machine customer service randomly pushes the answer A1 to the user, and the user continues to ask the next question;

[0119] IV. Determine the user's behavior trajectory: the user has not transferred to the manual customer service nor ended the session; then the weight of answer A1 is increased by 1;

[0120] Based on this contact between the user and the machine customer service, the following data will be formed:

[0121] Q: Which courier is it? Weights

[0122] A1: Express is SF Express 1

[0123] A2...

Embodiment 2

[0125] Embodiment 2: The user consults the human customer service to obtain the target answer

[0126] I. After the machine customer service replies to the user with A1, the user chooses to switch to manual customer service.

[0127] II. The user continues to ask "how many days will the logistics arrive, has it been 2 days?", the artificial customer service searches the candidate answer A3 from the knowledge base, and conducts secondary editing, and the answer obtained after editing is A4. After replying to the user with the answer A4, the user ends the session and gives a favorable comment; combined with Table 1, the weight of the positive factor corresponding to A4 is determined to be 2.

[0128] In such a case, for answer A1, the weight of the negative factor is determined to be 1 according to the way of obtaining the answer and the behavior trajectory (here, refer to "Ask questions to the machine customer service to obtain the target answer in Table 1, the machine customer...

Embodiment 3

[0141] Example 3: Data filtering

[0142] I. After running for a period of time, the entries in the knowledge base continue to increase, and there are multiple candidate answers for many questions. Suppose the following question-answer pairs exist:

[0143] Q: How many days can the logistics arrive? Weights

[0144] A1: Usually 3-5 days. 1 / 100

[0145] A2: Usually 3-5 days, please wait patiently. 99 / 100

[0146] II. There is a preset threshold stored in the knowledge base, such as 5 / 100. If the candidate answer whose weight is less than or equal to 5 / 100 is determined to be logically deleted, when the user asks the question "how many days will the logistics arrive?", only A2 will appear in the search result One candidate answer, A1 is filtered. The machine customer service can directly use the answer A2 to reply to the user, that is, the weight does not work when there is only one answer to the question. The question-answer pairs after data filtering are as follows:

...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a method for determining the weight of an answer, which includes: obtaining a behavior track; the behavior track represents a user's satisfaction degree to a target answer provided to the user; and determining the weight of the target answer according to the behavior track. The invention also discloses an answer weight determination device, a storage medium, an answer determination method, a device and a storage medium.

Description

technical field [0001] The present invention relates to computer technology, in particular to a method for determining the weight of an answer, a method for determining an answer, a device and a storage medium. Background technique [0002] At present, most of the intelligent customer service stores the required data in the knowledge base in advance. When the intelligent customer service or manual customer service is working, they can read the data stored in the knowledge base at any time. The data is composed of a large number of "question-answer pairs". Questions and corresponding answers are matched into groups. These "question-answer pairs" come from the usual communication between visitors and human customer service, or other dialogue occasions, and are then sorted out manually to form a knowledge base that can be used by intelligent customer service. [0003] Since the same question can be expressed in many ways, there may be multiple answers to the same similar quest...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/9032G06Q30/02
Inventor 仲潇
Owner MIGU CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap