Learning to rank (L2R) has emerged as a promising approach in handling the existing challenges of the Web search engines. However, there are major drawbacks with the present learning to rank techniques. Current L2R algorithms do not pay attention to the records of users’ interactions with Web search engines during their search sessions. Also, the learning process needs a large volume of features required from the users’ queries as well as the Web documents. Such situation has made the application of L2R techniques questionable in the real-world applications. Recently, based on the click-through data model and by generating click-through features, a novel approach is proposed, named as MGP-Rank. By the use of the layered genetic programming model, the MGP-Rank algorithm has achieved noticeable performance on the ranking of the English Web data. In this study, with respect to the specific characteristics of the Persian language, some suitable scenarios are presented for the generation of the click-through features. In this way, a customized version of the MGP-rank is proposed of the Persian Web retrieval. The evaluation results of this algorithm on the dotIR dataset, indicate its considerable improvement in comparison with the major ranking methods. This performance improvement is particularly more noticeable in the top part of the search results list, which is most frequently visited by the Web users.
- حق عضویت دریافتی صرف حمایت از نشریات عضو و نگهداری، تکمیل و توسعه مگیران میشود.
- پرداخت حق اشتراک و دانلود مقالات اجازه بازنشر آن در سایر رسانههای چاپی و دیجیتال را به کاربر نمیدهد.