Ph.D Thesis description

We aim to develop an Arabic IR System including:

1. New Indexing tool: Using either stems or roots as index terms offered considerable performance to Arabic IR (Information Retrieval) systems compared to the use of surface word for indexing. Many comparative works tried to find out the best from these two indexing approaches but until then, no of the two methods widely overtook the other. Each of the two index types performed better under different test circumstances in terms of recall and precision. So, our approach aims rather to combine the two methods in a way we take the advantages from both of them and try with this hybrid indexing to overcome their shortcomings.
2. Meta-search engine: Using a search engine doesn’t usually give good results in terms of recall and precision. This is due to coverage and scalability problems. For this, we aim to develop a meta-search engine.
3. Web Crawler: Web exploring became a real challenge because of the amount of web content is increasing and changing continuously. That’s why we need to develop a crawling tool in order to overcome recall and scalability problems.
4. Profile-based query reformulation tool: Query reformulation is one of the most important approaches to improve recall and precision results in IR system. In order to better get the user intention, we suggest reformulating queries based on his profile.