Introduction: ECAI 2016 is the best place for European AI scientific achievements. The conference provides researchers with a good opportunity to introduce and listen to the best research results of contemporary artificial intelligence.
Cluster-Driven Model for Improved Word and Text Embedding (Cluster-Driven Model for Improved Word and Text Embedding)
Abstract: Most existing text embedding models can only consider the relationship between text and its context (for example, cross-target text). However, information that transcends similar semantics (the overall context) reflects a rich semantic meaning, which is often overlooked. In this article, we propose a generic framework that uses global information to learn words and text representations. Our model can be easily integrated into existing local word embedding models to introduce different levels of global information for different downstream tasks. In addition, we look at our model from the perspective of a co-occurrence matrix, based on which a new weighted word-document matrix is ​​factorized to generate a textual representation. We conducted a series of experiments to evaluate the words and text representations learned through our model. The experimental results show that our model is superior or comparable to the best performance model.
The original program for this paper is at https://github.com/zhezhaoa/cluster-driven
First author introduction
Zhe Zhao
Renmin University of China School of Information
Via:ECAI 2016
PS : This article was compiled by Lei Feng Network (search "Lei Feng Network" public number) and it was compiled without permission.
Original paper download
