期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：4
页码：186-190
语种：English
出版社：Ayushmaan Technologies
摘要：Building a ranking model for each domain is a laborious and time consuming task which directly affect the search process in the web. In this paper, we are proposing a regularization based algorithm called ranking adaptation SVM (RA-SVM), to overcome this difficulties. This allows us to adapt an existing ranking model to a new domain, so that the amount of labeled data and the training cost is reduced and retain the performance as desired. This algorithm requires only the prediction from the existing ranking models it does not require the internal representations. Experiments performed over different dataset to demonstrate the application of our method and its performance.
关键词：Information Retrieval;Support Vector Machines;Learning to Rank;Domain Adaptation
