Unlock instant, AI-driven research and patent intelligence for your innovation.

Method and device for generating legal answers

A technology for laws and regulations, applied in the field of legal answer generation methods and devices, can solve problems such as limited answer, long structure relationship, inappropriateness, etc., to achieve the effect of improving efficiency

Active Publication Date: 2022-01-14
河北省讯飞人工智能研究院 +2
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since this method is based on historical information retrieval, it relies too much on historical legal question and answer data, and cannot create answers that are not in the database, so the answers that can be provided are limited and the scope of application is narrow
[0003] It is impossible to semantically guarantee the pertinence and accuracy of answers or structurally guarantee the matching degree of questions and answers, and it is also impossible to create answers that are not in the database
In addition, most of the above databases are modeled with triples as the basic unit, and triples are only suitable for knowledge types that state facts, and are not suitable for legal clauses and regulations that are long in length and have complex structural relationships, so they are not suitable for application in legal quiz
In addition, users often need to spend more time selecting answers from lawyers on the consulting platform and answers recommended by intelligent consulting functions.

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 and device for generating legal answers
  • Method and device for generating legal answers
  • Method and device for generating legal answers

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0109] Question entered: "What is the appeal process?";

[0110] Entered answer: "The process of appeal is: 1..., 2..., 3...";

[0111] The output of the Q&A matching model: 0.98.

example 2

[0113] Question entered: "What is the appeal process?";

[0114] Entered answer: "Hello, it is recommended to call the police";

[0115] The output of the Q&A matching model: 0.20.

[0116] In the method provided by the embodiment of the present invention, the loss function value can be finally obtained based on the matching degree, and the preset parameters in the initial model are iteratively optimized based on the loss function value to obtain the reinforcement learning model, so that the reinforcement learning model can output the answer data , the output results are more accurate, professional and detailed.

[0117] Based on the content of the foregoing embodiment, as an optional embodiment, the preset condition is that the loss function value of the iterative optimization process converges. Specifically, for the nth iterative optimization process, the convergence of the loss function value of the nth iterative optimization process may refer to the loss function value o...

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

Embodiments of the present invention provide a method and device for generating legal answers, which belong to the technical field of natural language processing. Including: obtaining question data, which is related to legal consultation; inputting question data into the reinforcement learning model, and outputting answer data. Since the reinforcement learning model is trained based on the initial model, and the type of the initial model is a codec model, rather than based on historical information retrieval, it does not need to rely on historical legal question and answer data, and can create new answer data for legal questions . Therefore, it is possible to provide answers covering a wide range. Secondly, because the type of the initial model is a codec model, there is no need to represent the training data in the form of triples, and the reinforcement learning model trained by the initial model is more suitable for answering legal questions.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of natural language processing, and in particular, to a method and apparatus for generating a legal answer. Background technique [0002] With the popularization of legal knowledge, more and more people tend to use legal means to safeguard their own rights and interests, and to solve problems fairly and justly through legal means. However, the resolution rate of legal problems is far lower than the occurrence rate of legal problems. Due to the rising demand for legal consultation, there are more and more online legal consultation platforms and platform-registered lawyers, but it is difficult for users to obtain professional and detailed legal help from legal consultation platforms. Most platform-registered lawyers tend to provide short and general answers and induce The questioner conducts a paid consultation or turns to an offline consultation. In recent years, some legal consultatio...

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): G06F16/332G06F16/33G06F40/295G06Q50/18
CPCG06F16/3329G06F16/3344G06Q50/18G06F40/295
Inventor 王梓玥王宝鑫伍大勇王士进胡国平刘挺
Owner 河北省讯飞人工智能研究院