Entity Extraction and Resolution for Legal Document Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems, such as West LiveNote, while powerful for legal research and document management, are not effectively accessible in real-time settings like depositions and trials, and struggle with automatically discerning relational and event information from vast electronic documents.
Innovation Solution
The development of a computer-implemented method and system for named-entity tagging, resolving, and event extraction, which includes an entity tagger, resolver, and relationship extractor, integrated with a Litigation Support System (LSS) to identify and link entities and events within documents, enabling efficient data extraction and presentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review of deposition transcripts and documents is performed, then accuracy in identifying entities and relationships is improved, but time consumption and productivity deteriorate
Solution Approach 1:
The system performs automatic entity detection, relationship extraction, and event identification without requiring manual review. The computer-implemented method autonomously processes deposition transcripts and legal documents to identify entities, relationships, and events, eliminating the need for attorneys and paralegals to manually examine thousands of pages of documents while maintaining high accuracy through machine learning algorithms
Solution Approach 2:
The patent replaces the mechanical process of manual document review with an automated computational system. The entity detection module, relationship extraction module, and event identification module work together to automatically process and analyze legal documents, substituting human manual labor with computer-based automated processing that achieves both high accuracy and high productivity
2Loss of information
If comprehensive entity detection and relationship extraction is implemented, then information completeness is improved, but system complexity increases
Solution Approach 1:
The system is divided into distinct functional modules: an entity detection module that identifies entities in text, a relationship extraction module that determines relationships between entities, and an event identification module that detects events. This segmentation allows each module to specialize in specific tasks, achieving comprehensive information extraction while managing system complexity through modular design
Solution Approach 2:
The system is designed to handle multiple types of legal documents (deposition transcripts, pleadings, motions, judgments) and extract multiple types of information (entities, relationships, events) using a unified computational framework. This multi-functionality approach ensures comprehensive information extraction across diverse document types without requiring separate specialized systems for each document type
3Ease of operation
If real-time access to research resources is provided during depositions and trials, then ease of operation is improved, but system adaptability requirements increase
Solution Approach 1:
The system pre-processes and indexes legal documents, entities, relationships, and events before deposition or trial activities. This preliminary processing creates a ready-to-query knowledge base that can be rapidly accessed in real-time during depositions and trials, providing ease of operation while the system adapts to different legal proceedings through the pre-established comprehensive data structure
Data Source
AI summary
For automated text processing, the inventors devised, among other things, an exemplary system that includes an entity tagger, an entity resolver, a text segment classifier, and a relationship extractor. The entity tagger receives an input text segment, and tags named entities with the segment as being a person, company, or place. The entity resolver accesses authority files, and associates the persons and companies named in the text segment with specific entries in the files. The text segment classifier determines whether the text segment includes a relationship event, such as job-change event or merger and acquisition event, and if an event is detected, the relationship extractor determines the event role of entities named in the segment. For example, the extractor determines for a merger and acquisition event, which named company was the acquirer and which was acquired.