Information Extraction for Freight-Related Natural Language Queries

TL;DRAbstract

The ability to retrieve accurate information from databases without an extensive knowledge of the contents and organization of each database is extremely beneficial to the dissemination and utilization of freight data. Advances in the artificial intelligence and information sciences provide an opportunity to develop query capturing algorithms to retrieve relevant keywords from freight-related natural language queries. The challenge is correctly identifying and classifying these keywords. On their own, current natural language processing algorithms are insufficient in performing this task for freight-related queries. High performance machine learning algorithms also require an annotated corpus of named entities which currently does not exist in the freight domain. This paper proposes a hybrid named entity recognition approach which draws on the individual strengths of models to correctly identify entities. The hybrid approach resulted in a greater precision for named entity recognition

Chat with Paper

AI Agents for this Paper