Scholarlink

Reading improves the reader’s vocabulary and knowledge of the world. It can open minds to different ideas which may challenge the reader to view things in a different light. Reading books benefits both physical and mental health of the reader, and those benefits can last a lifetime. It begins in early childhood and continue through the senior years. A good book should make the reader curious to learn more, and excited to share with others. For some readers, their reluctance to read is due to competing interests such as sports. For others, it is because reading is difficult and they associate it with frustration and strain. A lack of imagination can turn reading into a rather boring activity. In order to encourage adults to read, we propose an elegant book recommender for adults based on either the Vector Space model or a machine-learning model, both are content-based approaches without utilizing the actual content of a book, which is often unavailable due to the copyright constraint. The Vector Space model is simply an algebraic model for representing text documents as vectors of identifiers, whereas the machine-learning model solely relies on a trained model using a pre-defined labeled item set. With that, we offer the user the choice between an algebraic model or a machine learning model to be considered. In order to analyze the performance of the proposed book recommender, BkRec, we have included in the empirical study an online evaluation to verify the novelty of BkRec in recommending books to adults, besides conducting an offline evaluation. The online evaluation determines whether recommended book are perceived as preferable by ordinary users, i.e., adults, which offers another perspective on the performance of the recommender. This evaluation is based on real users’ assessments of BkRec which goes beyond the offline performance analysis conducted. In the online performance evaluation, we determine the ideal number of appraisers and test cases extracted from the BookCrossing dataset for the evaluation of BkRec to further confirm the reliable and objectiveness of BkRec. Experimental results have verified the effectiveness of BkRec.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Personalized Book Recommendations for Adults Using Deep Learning and Sophisticated Filtering Approaches

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yiu-Kai Ng

摘要

Reading improves the reader’s vocabulary and knowledge of the world. It can open minds to different ideas which may challenge the reader to view things in a different light. Reading books benefits both physical and mental health of the reader, and those benefits can last a lifetime. It begins in early childhood and continue through the senior years. A good book should make the reader curious to learn more, and excited to share with others. For some readers, their reluctance to read is due to competing interests such as sports. For others, it is because reading is difficult and they associate it with frustration and strain. A lack of imagination can turn reading into a rather boring activity. In order to encourage adults to read, we propose an elegant book recommender for adults based on either the Vector Space model or a machine-learning model, both are content-based approaches without utilizing the actual content of a book, which is often unavailable due to the copyright constraint. The Vector Space model is simply an algebraic model for representing text documents as vectors of identifiers, whereas the machine-learning model solely relies on a trained model using a pre-defined labeled item set. With that, we offer the user the choice between an algebraic model or a machine learning model to be considered. In order to analyze the performance of the proposed book recommender, BkRec, we have included in the empirical study an online evaluation to verify the novelty of BkRec in recommending books to adults, besides conducting an offline evaluation. The online evaluation determines whether recommended book are perceived as preferable by ordinary users, i.e., adults, which offers another perspective on the performance of the recommender. This evaluation is based on real users’ assessments of BkRec which goes beyond the offline performance analysis conducted. In the online performance evaluation, we determine the ideal number of appraisers and test cases extracted from the BookCrossing dataset for the evaluation of BkRec to further confirm the reliable and objectiveness of BkRec. Experimental results have verified the effectiveness of BkRec.