Developed NLP Techniques for Entity and Relation Extraction

1 min read

Developed NLP Techniques for Entity and Relation Extraction

  • Day: 2024-09-11
  • Time: 22:10 to 23:20
  • Project: Dev
  • Workspace: WP 2: Operational
  • Status: In Progress
  • Priority: MEDIUM
  • Assignee: Matías Nehuen Iglesias
  • Tags: NLP, Entity Extraction, Relation Extraction, Python, Spacy

Description

Session Goal: The session aimed to enhance natural language processing (NLP) techniques for entity and relation extraction, focusing on improving parsing and extraction in complex legal and structured texts.

Key Activities:

  • Developed a Python function using spaCy and networkx for entity and relationship extraction, creating graph structures from sentences.
  • Improved entity and relation extraction in legal texts by addressing conjunctions, indirect objects, and complex sentence structures.
  • Enhanced sentence parsing for better recognition of subjects, verbs, and objects, focusing on actions involving persons and institutions.

Achievements:

  • Successfully implemented a structured approach for building a knowledge graph from natural language summaries.
  • Improved the accuracy of relation extraction functions by refining parsing techniques and updating Python code.

Pending Tasks:

  • Further testing and validation of the improved NLP techniques on diverse datasets to ensure robustness and accuracy.
  • Integration of the developed functions into larger NLP workflows for comprehensive text analysis.

Evidence

  • source_file=2024-09-11.sessions.jsonl, line_number=2, event_count=0, session_id=9006a2bcf61f60daf2a41ed510371c51ed9b4d0a2af729af096c7cec6c068f6c
  • event_ids: []

Graph View

Backlinks