Thesis (M.Sc.) - Cairo University - Faculty of Computer and Information - Department of Computer Science

Anaphora resolution is one of the challenging tasks of natural language processing. It is very important since without it a text would not be fully and correctly understood, and without finding the proper antecedent, the meaning and the role of the anaphor cannot be realized. Information about anaphoric relations has a significant role in applications such as automatic summarization, machine translation, Question-Answering, and others. The motivation behind this work is to produce Arabic corpus annotated with Anaphoric information, which is a valuable resource that enables future research in Arabic anaphora resolution, and helps in future work in analyzing Quran script. It will be available resource used for training and testing anaphora resolution systems, and evaluating and optimizing existing approaches of anaphora resolution. My own motivation for choosing Arabic language is that it is less-studied language and works on Arabic anaphora resolution are very few. By providing a representative corpus it will benefit researchers, encourage work in Arabic anaphora resolution, and help in extracting empirical patterns and rules for either building new anaphora resolution approaches or other Arabic NLP applications. The provided multi-dimensional feature vector will provide anaphoric information needed to build statistical anaphora resolution systems

Issued also as CD