Method for finding lawyers on line

Abstract

The invention belongs to the technical field of computer networks, and provides a method for finding lawyers online. In order to solve the problem that a traditional lawyer finding mode is inaccurate in matching, case description and files are received through a user input module, and preprocessing, semantic understanding and key information extraction are carried out; the dialogue interaction module obtains legal service demand information through multiple rounds of interaction; the law office-lawyer-judgment document database stores lawyer agent historical case information; the lawyer recommendation module combines various methods to calculate similarity scores, screens out judgment documents with good case results, recommends lawyers, and displays information such as lawyer agent fees; and the charging mode management module realizes the unified setting of the charging function. According to the method, various artificial intelligence technologies are comprehensively applied, efficient and accurate lawyer recommendation service is provided for the user, the matching efficiency is improved, and the transparency and reliability of the recommendation process are enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of computer networks, and particularly relates to a method for finding a lawyer online. Background Art

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

[0003] The present invention aims to provide a method for finding a lawyer online to help users find highly cost-effective lawyers suitable for their needs.

[0004] System Composition: User Input Module Receives the user's description of the case situation and uploaded files, preprocesses and semantically understands the case description input by the user, and extracts key information from the uploaded files through OCR and named entity recognition.

[0005] Dialogue Interaction Module Obtains user legal service requirement information through multi-round voice or text dialogue interaction, uses a pre-trained language model to perform intent recognition and entity extraction on user input, including information such as case type, case situation, dispute, evidence, lawyer fee payment mode and budget, jurisdiction court, etc., to construct user case information; dynamic questioning strategy: when key case information is missing, trigger questioning based on the principle of maximizing information gain; supports multi-modal input, such as extracting key information from evidence files uploaded by users through OCR and named entity recognition (NER), etc.

[0006] Law Firm-Lawyer-Judgment Document Database The Law Firm-Lawyer-Judgment Document Database contains judgment documents of lawyers' historical cases (anonymized judgment documents, judgment document summaries, and other judgment document variants). Information such as lawyer name, practicing law firm name, law firm location, legal field, litigation request, dispute, evidence, case situation, case result label, accepting court, lawyer fee charging mode and parameters, etc.

[0007] Lawyer Recommendation Module Combining the information of the user's description and the uploaded file, one or more of the following methods are adopted: traditional statistical methods such as TF-IDF (Term Frequency-Inverse Document Frequency), BM25 (Best Matching 25), Jaccard similarity, cosine similarity, etc.; semantic and vector space methods such as Word2Vec, GloVe, FastText, Doc2Vec, etc.; topic models such as LDA (Latent Dirichlet Allocation), LSI (Latent Semantic Indexing), etc.; deep learning methods such as Sentence-BERT, SimCSE, DSSM (Deep Structured Semantic Model), Transformer-based Models (such as BERT, GPT, T5, BLOOM, etc.); fast retrieval and optimization techniques such as approximate nearest neighbor (ANN) search, locality-sensitive hashing (LSH), etc.; hybrid methods such as multimodal fusion, graph embedding (such as Node2Vec, GraphSAGE), etc. Calculate the similarity score between the user input and the judgment documents, screen out the judgment documents with high similarity from the judgment documents with good case results in the law firm-lawyer-judgment document database, recommend multiple lawyers to the user and display the lawyer's attorney fees for this case according to the attorney fee charging function set by the law firm. The lawyer information displayed to the user includes: lawyer's name, practice number, practicing law firm, similar judgment documents (anonymized judgment documents and their abstracts), attorney fees for this case, years of practice, lawyer's photo, etc.

[0008] Fee Model Management Module The fee model management module allows law firms to set different charging functions according to information such as case types and case subject amount ranges, and unifies the eight attorney fee charging methods in the current market. When the system obtains information such as the user's case type and case subject amount, it can automatically calculate the attorney fees for this case and display them in the lawyer recommendation, showing the case attorney fees to the user, facilitating the user to compare prices among lawyers at the same level, and optimizing the user's decision-making.

[0009] Method Steps Receive user input Receive the user's description of the case and the uploaded file, preprocess and semantically understand the user's input, and extract key information from the uploaded file through OCR and named entity recognition. Or obtain the user's legal service demand information through multi-round voice or text dialogue interaction.

[0010] Lawyer Recommendation According to the user input, screen out the judgment documents with high similarity from the judgment documents with good case results in the law firm-lawyer-judgment document database, and recommend lawyers to the user Description of the Drawings

[0011] The accompanying drawings are used to provide a further understanding of the present invention and form a part of the description. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. Figure 1 It is a schematic diagram of a method flow for finding a lawyer online.

[0012] Example User A (the plaintiff) described the course of events in multiple rounds and uploaded screenshots of online shopping orders. The system parsed the user's description to extract key information. The user's description and evidence files were compared with all the judgments of law firms with good case results in the jurisdiction of the court where the case is located in the law firm-lawyer-judgment document database to find judgments and lawyers with relatively high similarity. Finally, the system recommended multiple eligible lawyers to the user and displayed the agency fees of the lawyers for this case for the user to choose.

[0013] User B (the defendant) uploaded the plaintiff's complaint served to him by the court to the system after anonymization; the system parsed and extracted key information. The plaintiff's complaint was compared with all the judgments of law firms with good case results in the jurisdiction of the court where the case is located in the law firm-lawyer-judgment document database to find judgments and lawyers with relatively high similarity. Finally, the system recommended multiple eligible lawyers to the user and displayed the agency fees of the lawyers for this case for the user to choose.

[0014] Conclusion By comprehensively using a variety of artificial intelligence technologies and combining user descriptions, evidence files, law firm-lawyer-judgment document databases, fee information, and case location information, the present invention provides users with an efficient and accurate lawyer recommendation service. 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 online, characterized in that: The invention comprises: The user input or dialogue interaction module pre-processes and semantically understands the user's description of the case, and extracts key information from the uploaded file through OCR and named entity recognition. It is more suitable for users who are defendants. The dialogue interaction module obtains user legal service demand information through multiple rounds of voice or text dialogue interaction, and is more suitable for users who are plaintiffs. Law firm-lawyer-judgment document database, including the judgment documents of the lawyer's historical cases, the lawyer's name, the name of the law firm, the location of the law firm, the legal field, the litigation request, the dispute, the evidence, the case situation, the case result label, the court accepting the case, the lawyer's agency fee charging model and parameters, etc. The lawyer recommendation module combines the information of the user description and uploaded files and adopts one or more of the following methods: traditional statistical methods, semantic and vector space methods, topic models, deep learning methods, fast retrieval and optimization techniques, hybrid methods or other related algorithms; calculates the similarity score between the user input and the judgment document, selects the judgment documents with high similarity from the judgment documents with good case results in the law firm-lawyer-judgment document database, recommends multiple lawyers to the user and displays the lawyer's agency fee for the case according to the lawyer's agency fee charging function set by the law firm.

2. The system according to claim 1, characterized in that The dialogue interaction module includes: Use pre-trained language models to identify user input intent and extract entities to build user case information; dynamic questioning strategy: when key case information is missing, trigger questioning based on the principle of maximizing information gain; support multimodal input, extract key information from user-uploaded files through OCR and named entity recognition (NER).

3. The system according to claim 1, characterized in that The judicial documents include: anonymized judicial documents, judicial document summaries and other judicial document variants.

4. The system according to claim 1, characterized in that The similarity calculation adopts one or more of the following technologies: traditional statistical methods, such as TF-IDF (term frequency-inverse document frequency), BM25 (Best Matching 25), Jaccard similarity, cosine similarity, etc.; semantic and vector space methods, such as Word2Vec, GloVe, FastText, Doc2Vec, etc.; topic models, such as LDA (latent Dirichlet allocation), LSI (latent semantic indexing), etc.; deep learning methods such as Sentence-BERT, SimCSE, DSSM (deep structured semantic model), Transformer-basedModels (such as BERT, GPT, T5, BLOOM, etc.), etc.; fast retrieval and optimization technologies, such as approximate nearest neighbor (ANN) search, local sensitive hashing (LSH), etc.; hybrid methods, such as multimodal fusion, graph embedding (such as Node2Vec, GraphSAGE), etc.

5. The system according to claim 1, characterized in that The charging model and parameters in the law firm-lawyer-judgment document database and the lawyer recommendation module show that the lawyer's agency fee for the case is: the law firm sets different charging functions based on information such as the case type and the case amount range. When the system obtains information such as the user's case type and case amount, it can automatically calculate the lawyer's agency fee for the case and display it in the lawyer recommendation.

6. The system according to claim 1, characterized in that The lawyer information recommended to users in the lawyer recommendation module includes: lawyer's name, practice number, practicing law firm, similar judgment documents (anonymized judgment documents and their summaries), lawyer's agency fees for the case, years of practice, lawyer's photo and other information.

7. 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.

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

Images