Bootstrapping the Albanian Information Retrieval

摘要

In this paper we investigate the Albanian language and try to uncover the characteristics of the language that will permit the Information Retrieval (IR) community to develop IR systems adapted for the specific language. As a consequence of our study (investigation) we provide a naive-single-step (rudimentary) stemming algorithm for the Albanian language. A stopword list is also created. Human experts are contacted for the evaluation of the provided stemming algorithm. The evaluation method used and the observation of the method's results uncover more rules, which could improve the capabilities of the rudimentary stemming algorithm. We believe that our approach for this specific language could become a standard way for building Information Retrieval functionalities (tools, functions, etc) for languages less perused, as is the language studied in this paper.