Legal consultation evaluation system based on scenarization

Abstract

The invention discloses a legal consultation evaluation system based on scenarization, which converts a legal consultation evaluation process into spoken language case description by combining elementcarding and recognition under a dispute type, and supports input in various forms in combination with a question and answer supplement form. Secondly, after spoken description of the user, automatically identifying a dispute type and an appeal type through machine learning and deep learning technologies in combination with manual annotation data; thirdly, in combination with a scene positioning engine, the input of the user is positioned to a specific scene under the appeal, so that a premise is provided for subsequent targeted questioning and generation of a consultation evaluation report; according to different user inputs, the positioned specific scene may be one, and may also be a few of most feasible scenes; therefore, in combination with a dynamic multi-round interactive question and answer mode, the scene of the user is determined in a least question mode through targeted questions; and generating a targeted legal consultation evaluation report according to the determined scene.

Description

technical field [0001] The invention relates to the technical field of legal consultation, in particular to a context-based legal consultation evaluation system. Background technique [0002] There are tens of millions of civil cases and even more legal consultations every year. Faced with the increasing demands of citizens for legal services, various entities have launched corresponding legal service platforms or tools; massive judgment documents and public legal data have made the application of various legal services possible. [0003] The current legal consultation and evaluation system is mainly realized by users selecting the type of questions and asking a large number of fixed questions. Taking professional consultation and evaluation in the civil field as an example, it includes the following steps: [0004] 1) The user selects a specific question type—here, by selecting a question type, the scope of consultation is limited. A typical question type is a certain type...

AI Technical Summary

Problems solved by technology

[0009] 1. The questions that the system requires users to answer are basically fixed and have nothing to do with the user's specific situation; for the same category of questions, everyone is asked the same question, that is, they cannot ask targeted questions;

[0010] 2. The system needs to ask the user a lot of questions. Since the system does not know which specific scenario the user belongs to, a large amount of user information needs to be collected, resulting in a complicated, time-consuming consultation and evaluation process, and a low completion rate for users;

[0011] 3. The consultation and evaluation reports provided are usually filled in templates, which contain a lot of general and fixed information, and cannot draw a clear and directional conclusion for specific scenarios of users;

[0012] 4. Users must correctly select a specific type of dispute before they can start systematic consultation and evaluation; due to the lack of professional legal knowledge of most users and the complexity of the legal knowledge system, it is difficult for users to choose or easy to make wrong choices, such as In medical scenarios, when users go to the hospital to see a doctor, they are used to guiding them to the correct department through a guidance desk. Similar functions are also required in legal service scenarios;

[0013] 5. The system cannot support users' oral descriptions. Users can only passively reply to questions, but cannot specifically introduce their own situation, resulting in less effective user input and poor experience, and the system is too mechanical. their own case, and then they can combine professional knowledge to ask corresponding questions to further confirm the case) Contradictory

Images