Corpus-Based Genre Analysis: Computer Science Research Article Introductions Conventional definitions of genres, based on the notions of specific conventions such as of content (theme, setting etc.) and form (structure and style) have been disputed. Some scholars do not believe in the rigid rules of inclusion and exclusion of texts in a particular genre as they can be recognized intutively as instances of repetition and difference because of their 'family resemblences' among texts. Swales (1990) prefers the psycholinguistic concept of `prototypicality'. Genres usually go through phases or cycles of popularity as the crucial ideological concerns of the time in which they are popular are embodied in the generic conventions. The popular genre of research article, amongst the research and academic community, is undergoing a continuous evolution. Many scholars have attempted to explore this complex process of writing research article. The list is long, to name some: Berkenkotter and Huckin (1995); Montgomery (1996); Salager-Meyer (1998); Atkinson (1999); Valle (1999); Gross et al. (2002). The work of these scholars includes research articles from different disciplines. However, such scholarly work in the field of Computer science is limited. Cooper (1985), Posteguille (1999) and Anthony (1999) studies are either too broad or too narrow. As compared to these works, the present study addresses the issue at a greater length and is more in depth. The increasing use of computer led text corpora containing millions of words inspired the utilization of the corpus-based techniques for the present research. A corpus of 56 Research articles was created electronically. These articles were taken from five different journals of IEEE, the leading Computer society of the world. Wordsmith tools such as word frequency list, key word, collocation and concordance were applied to the corpus. Secondly, Swales (2004) CARS model was applied for the rhetorical analysis. Lexico-erammaticl analysis was done in terms of the rhetorical objectives of writing Introductions. The findings of the research as discussed in Chapter four focus on the syntactic and lexical patterns evident in the data. Theses include interesting N-grams (three and four word clusters); voice of the author (very different from the authors of other disciplines) and passivization of verbs. These stylistic excursions make an initial contribution to our understanding of Computer science research articles Introductions. The last three chapters of the dissertation constitute the heart of the discoursal analysis of the 56 Introductions in the corpus. These examine the structural-rhetorical features of the moves and steps involved, and the possible linkages between form and function. A revised CARS model has been suggested for writing Introductions of computer science research articles. Some recommendations have been put forward. The dissertation concludes with a note on the pedagogical relevance of the study.