Enhancing Collaborative Variational Autoencoder with Tag and Citation Information for Scientific Article Recommendation
Enhancing Collaborative Variational Autoencoder with Tag and Citation Information for Scientific Article Recommendation
复制标题
DOI：
10.5109/2230667
复制
发表时间：
2018-11
期刊：
影响因子：
--
通讯作者：
Xanh Ho;Akiko Aizawa
中科院分区：
文献类型：
--
作者：
Xanh Ho;Akiko Aizawa
协作深度学习（CDL）和协作变分自动编码器（CVAE）等混合方法已成为科学文章推荐系统中最先进的方法。然而，它们通常只使用文章标题和摘要中的信息，而忽略了标签和引文中潜在的有用信息。因此，他们可能会错过包含与其他文章截然不同的内容的文章，尽管这些文章呈现相同的主题。我们通过开发CiT-CVAE模型来解决这个问题，该模型在提供推荐时考虑标签和引用信息。实验结果表明，与CDL和CVAE相比，该模型取得了一致的效果.
Hybrid methods such as collaborative deep learning (CDL) and collaborative variational autoencoder (CVAE) have become state-ofthe-art methods in recommender systems for scientific articles. However, they typically use only information from titles and abstracts of articles, and ignore potentially useful information in the tags and citations. Therefore, they may miss articles that contain vastly different content from other articles, although those articles present the same topic. We addressed this problem by developing the CiT-CVAE model that considers tag and citation information when providing recommendations. Our experimental results indicate that the proposed model achieves consistent improvement compared with CDL and CVAE.