A Lightweight Semantic Patent Search System for the Electric Vehicle Domain

Khurana, Priya Sethi (2025) A Lightweight Semantic Patent Search System for the Electric Vehicle Domain. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (989kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

In technology fields like Electric Vehicles (EVs), where innovation is fast-paced and ideas often intersect, finding truly relevant patents can be challenging. This study introduces a tailored patent similarity search system designed to address that challenge by focusing on the semantic content of patent claims rather than surface-level keywords. Using a compact transformer model (msmarco-MiniLM-L-6-v3), combined with a lightweight claim-weighting approach, the system captures both the meaning and importance of different claims. These are indexed in a fast, scalable vector search engine (FAISS) for efficient retrieval. The system is designed to be both practical and easy to scale, showing strong performance in identifying closely related patents while filtering out irrelevant results. A series of experiments validate its effectiveness across both real-world and synthetic patent queries, supporting its role as a useful tool for researchers and IP professionals in fast-evolving domains.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Anant, Aaloka UNSPECIFIED
Subjects:	T Technology > TL Motor vehicles. Aeronautics. Astronautics Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Motor Industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	24 Aug 2026 12:15
Last Modified:	24 Aug 2026 12:15
URI:	https://norma.ncirl.ie/id/eprint/9604

Actions (login required)

View Item