Anaphora Resolution (AR) has attracted the attention of many researchers because of its relevance to Machine Translation, Information Retrieval, Text Summarization and many other applications. AR is a complicated problem in NLP especially in Semitic languages because of their complex morphological structure. Anaphora can be defined as a linguistic relation between two textual entities which is determined when a textual entity (the anaphor) refers to another entity of the text which usually occurs before it (the antecedent). The process of determining the antecedent of an anaphor is referred to as anaphora resolution. In this chapter, we present an account of the anaphora resolution task. The chapter consists of ten sections. The first section is an introduction to the problem. In the second section, we present different types of anaphora. Section 3 discusses the determinants and factors to anaphora resolution and its effect on increasing system performance. In section 4, we discuss the process of anaphora resolution. In section 5 we present different approaches to resolving anaphora and we discuss previous work in the field. Section 6 discusses the recent work in anaphora resolution, and section 7 discusses an important aspect in the anaphora resolution process which is the evaluation of AR systems. In sections 8 and 9, we focus on the anaphora resolution in Semitic languages in particular and the difficulties and challenges facing researchers. Finally, section 10 presents a summary of the chapter.
|Anaphora Resolution||0 bytes|