https://scholars.lib.ntu.edu.tw/handle/123456789/488363
標題: | Less is More: Filtering Abnormal Dimensions in GloVe. | 作者: | Lee, Yang-Yin Ke, Hao Huang, Hen-Hsen HSIN-HSI CHEN |
關鍵字: | glove; semantic relatedness; word embedding | 公開日期: | 2016 | 起(迄)頁: | 71-72 | 來源出版物: | Proceedings of the 25th International Conference on World Wide Web, WWW 2016, Montreal, Canada, April 11-15, 2016, Companion Volume | 摘要: | GloVe, global vectors for word representation, performs well in some word analogy and semantic relatedness tasks. However, we find that some dimensions of the trained word embedding are abnormal. We verify our conjecture via removing these abnormal dimensions using Kolmogorov-Smimov test and experiment on several benchmark datasets for semantic relatedness measurement. The experimental results confirm our finding. Interestingly, some of the tasks outperform the state-of-the-art model SensEmbed by simply removing these abnormal dimensions. The novel rule of thumb technique which leads to better performance is expected to be useful in practice. © 2016 owner/author(s). |
URI: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85070777597&doi=10.1145%2f2872518.2889381&partnerID=40&md5=7ca9af444e67da95e7573c659556ae0a | DOI: | 10.1145/2872518.2889381 | SDG/關鍵字: | Embeddings; ART model; Benchmark datasets; Embeddings; Glove; Kolmogorov; Less is mores; Semantic relatedness; State of the art; Word embedding; Word representations; Semantics |
顯示於: | 資訊工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。