Cognitive location-aware information retrieval by agent-based semantic matching

Agents operating in both wired and wireless networks find and retrieve location-aware information. Agents in our system are required to endow with the full range of cognitive abilities, including perception, use of natural language, learning and the ability to understand the user query. The speed and accuracy of retrieval and the usefulness of the retrieved data depends on a number of factors including constant or frequent changes in its content or status, the effects of environmental factors such as the weather and traffic and the techniques that are used to categorize the relevance of the retrieved data. In this paper, we propose semantic TFIDF, an agent-based system for retrieving location-aware information that can improves the speed of retrieval while maintaining or even improving the accuracy by making use of semantic information in the data to develop smaller training sets. In our method, intelligent agents first gather location-aware data and then, using semantic graphs in the WordNet English dictionary, they classify, match and organize the information to find a best match for a user query. Our experiments compared our proposed system with three other commonly used systems and showed it to be significantly faster and more accurate.