Developing an Interactive Communicative Search Agent

Abstract

Developments in semantic web are changing how websites are written for the internet. The categorization principles are applied to closed systems such as a company's internal document database; however, these changes are not as quickly re ected in semantic search due to existing problems within the natural language programming eld. This project aims to develop a framework for a di erent type of search: one that combines linguistic principles and current search options and tailor returned results according to user input. The Python program- ming environment and Natural Language Toolkit libraries are the pri- mary resources used to develop this program. Once a basic working code is developed, search results are compared against an increasingly-popular alternative search tool. The results from these searches provide feedback on which areas the program needs to develop further.