Method for searching lawyers based on similarity comparison of user description, uploaded materials and discriminants

Abstract

The invention discloses a lawyer searching method based on user description and similarity comparison of uploaded materials and discriminants. The method is applied to the technical field of computer networks. Aiming at the limitation of a traditional lawyer finding mode, a service capable of accurately matching user requirements is provided. The system design comprises a user input module, a preprocessing module, a semantic understanding module, a discriminant database, a similarity calculation module, a charging mode management module and a lawyer recommendation module. According to the method, preprocessing, semantic understanding and feature extraction are carried out on description input by a user and files uploaded by the user, then similarity calculation is carried out on cases in a judgment database, and lawyers with similar case winning experience are recommended. In addition, the system can set different charging modes according to the case type, the case quota interval, the complaint rate and the like, and transparent charging is ensured. According to the method, the probability that the user finds a proper lawyer is effectively improved, and the transparency and reliability of the lawyer recommendation process are enhanced.

Description

Technical Field

[0001] The present invention belongs to the field of computer network technology, and in particular to a method for finding a case agent lawyer based on similarity comparison between user description and uploaded materials and a judgment. Background Art

[0002] Traditional ways of finding lawyers usually rely on personal recommendations or simple search functions on online platforms, which often fail to accurately match users' actual needs. The present invention provides a method that can display prices and accurately find lawyers with experience in winning similar cases. Summary of the invention

[0003] The present invention aims to provide a method for comparing the similarity between user descriptions and uploaded materials and judgments, thereby helping users find the winning lawyer that best suits their needs.

[0004] System composition: User Input Module Receive users' natural language descriptions and upload files (including pictures, videos, and audio).

[0005] Preprocessing module The natural language description input by the user is preprocessed by word segmentation, stop word removal, part-of-speech tagging, etc., and the uploaded evidence file is format converted, compressed, and feature extracted.

[0006] Semantic Understanding Module Use natural language processing technology to parse user descriptions, and use computer vision and audio analysis technology to identify and extract key information from evidence files. Including: Text analysis submodule: Use pre-trained language models (such as BERT, GPT, etc.) to perform deep semantic analysis on user descriptions; Image recognition submodule: Uses convolutional neural network (CNN) or Transformer architecture to perform object detection, scene classification, and text recognition on uploaded images; Video analysis submodule: Use 3D convolutional neural network (3D-CNN) or long short-term memory network (LSTM) to perform action recognition, scene segmentation and event extraction on uploaded videos; Audio analysis submodule: Use audio feature extraction technology (such as MFCC) to perform speech recognition and keyword extraction on the uploaded audio clips.

[0007] Judgment Database The judgment database contains a large number of historical judgment texts and their corresponding labels such as legal fields, litigation requests, case circumstances, disputes, lawyer information, case results, etc.

[0008] Similarity calculation module Combined with the information of the user description and the evidence file, a vector space model, a deep learning model or other related algorithms are used to calculate the similarity score between the input and the judgment text. One or more of the following technologies can be used: a similarity calculation method based on keyword matching; a similarity calculation method based on a vector space model; a similarity calculation method based on deep learning; a similarity calculation method based on multimodal fusion.

[0009] Charging model management module The charging model management module allows law firms to set different charging models and parameters based on case type, case amount range, and case win rate, and displays case agency fees to users to ensure transparent charging.

[0010] Lawyer recommendation module The lawyer recommendation module selects the winning lawyers from the judgment database based on the similarity score and recommends them to users based on the location of the case and the agency fee charging model. It includes a lawyer screening submodule that selects a list of qualified lawyer candidates from the judgment database based on the similarity score; The win rate assessment submodule evaluates candidate lawyers based on their win rate in similar cases and other performance indicators; The charging model matching submodule selects lawyers who meet the user's expected charging model based on the user's budget and needs; The lawyer ranking submodule comprehensively considers factors such as lawyers' professional experience, win rate in similar cases, user evaluation, charging model and fees, location of the law firm, etc., ranks candidate lawyers, and recommends top-ranked lawyers to users.

[0011] Method steps Receiving User Input Receive users' natural language descriptions and upload files (including pictures, videos, and audio).

[0012] Preprocessing Preprocess the natural language description input by the user, and perform format conversion, compression processing, and feature extraction on the uploaded evidence file.

[0013] Semantic Understanding Use natural language processing technology to parse user descriptions, and use computer vision and audio analysis technology to identify and extract key information from evidence files.

[0014] Search judgment Relevant judgment texts are retrieved from the judgment database and subjected to the same preprocessing and semantic understanding operations.

[0015] Similarity calculation Calculate the similarity score between the user description and evidence documents and the judgment text.

[0016] Charging mode settings Law firms are allowed to set different charging models and parameters based on case type, case amount range, and case win rate, and display the case agency fee charging model and price to users.

[0017] Case location screening Based on the case location information provided by the user, practicing lawyers in the relevant area are screened out.

[0018] Lawyer recommendation The winning lawyers are selected from the judgment database based on the similarity score, and the most suitable lawyers are recommended to users based on the location of the case and the agency fee charging model. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the description. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. Figure 1 It is a schematic diagram of the process of a method for finding a lawyer based on the similarity comparison between the user's description and uploaded materials and the judgment.

[0020] Example of embodiment In the online sales contract case, user A described the incident and uploaded a screenshot of the online shopping order; The system parses user descriptions through a semantic understanding module and extracts key information through image recognition; The similarity calculation module compares the user description and evidence files with the online sales contract judgments in the judgment database to find the judgment with the highest similarity; The charging model management module selects lawyers who meet the requirements according to the charging model selected by the user; In the end, the system recommends several qualified lawyers to the user and provides detailed fee information for the user to choose.

[0021] in conclusion The present invention provides users with an efficient and accurate lawyer recommendation service by comprehensively utilizing various artificial intelligence technologies such as natural language processing, computer vision, audio analysis, etc., combined with user descriptions, evidence documents, judgment databases, charging models, and case location information. This not only increases the probability of users finding suitable lawyers, but also enhances the transparency and reliability of the lawyer recommendation process.

Claims

1. A method for finding a lawyer for a case based on a similarity comparison between a user's description and uploaded materials and a judgment, characterized in that: The invention comprises: A user input module, used to receive the user's natural language description and upload files; The preprocessing module performs preprocessing operations such as word segmentation, stop word removal, and part-of-speech tagging on the natural language description input by the user, and performs format conversion, compression processing, and feature extraction on the uploaded file; The semantic understanding module uses natural language processing technology to parse the content described by the user and, when necessary, uses computer vision and audio analysis technology to identify and extract key information from the file; Judgment database, which contains a large number of historical judgment texts and their corresponding labels such as legal fields, case circumstances, lawyer information, case results, etc. A similarity calculation module, which combines the information of the user description and the uploaded file and uses a vector space model, a deep learning model or other relevant algorithms to calculate the similarity score between the user input and the judgment text; The charging model management module allows law firms to set different charging models and parameters based on case types, case amount ranges, and case success rates, and displays case agency fees to users; The lawyer recommendation module selects winning lawyers from the judgment database based on similarity scores, and recommends them to users based on the location of the case and the agency fee charging model.

2. The system according to claim 1, characterized in that The semantic understanding module includes: The text analysis submodule uses pre-trained language models (such as BERT, GPT, etc.) to perform deep semantic analysis on user descriptions; The image recognition submodule uses a convolutional neural network (CNN) or Transformer architecture to perform object detection, scene classification, and text recognition on uploaded images; The video analysis submodule uses 3D convolutional neural networks (3D-CNN) or long short-term memory networks (LSTM) to perform action recognition, scene segmentation, and event extraction on uploaded videos; The audio analysis submodule uses audio feature extraction technology (such as MFCC) to perform speech recognition and keyword extraction on the uploaded audio clips.

3. The system according to claim 1 or 2, characterized in that: The similarity calculation module adopts one or more of the following technologies: The similarity calculation method based on keyword matching calculates the similarity by comparing the keywords appearing in the user description and the judgment; Based on the similarity calculation method of the vector space model, the user description and the judgment text are converted into vector representations, and the cosine similarity between them is calculated; A similarity calculation method based on deep learning uses a pre-trained neural network model to encode user descriptions and judgment texts and calculate the similarity score between them; A similarity calculation method based on multimodal fusion combines user descriptions with image, video, and audio information in evidence files for comprehensive similarity evaluation.

4. The system according to claim 1, characterized in that The lawyer recommendation module further includes: The lawyer screening submodule selects a list of qualified lawyer candidates from the judgment database based on the similarity score; The win rate assessment submodule evaluates candidate lawyers based on their win rate in similar cases and other performance indicators; The charging model matching submodule selects lawyers who meet the user's expected charging model based on the user's budget and needs; The lawyer ranking submodule comprehensively considers factors such as lawyers' professional experience, win rate in similar cases, user evaluation, charging model and fees, location of the law firm, etc., ranks candidate lawyers, and recommends top-ranked lawyers to users.

5. The system according to claim 1, characterized in that The charging model management module further includes: a charging model setting submodule, which allows law firms to flexibly set different charging models and their parameters (such as fixed fees, hourly billing, risk agency fees, etc.) according to factors such as case type, case winning rate, case amount range, and case complexity; a fee display submodule, which displays the case agency fee charging model and price to users to ensure transparent charging.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method steps described in any one of claims 1 to 4 are implemented.

7. An electronic device, comprising a processor and a memory, characterized in that: The memory stores a computer program, and when the computer program is executed by the processor, the method steps according to any one of the claims are implemented.

Images