In this paper, we propose a new methodology based on directed graphs and the TextRank algorithm to automatically induce general-specific noun relations from web corpora frequency c...
Traditional ranking mainly focuses on one type of data source, and effective modeling still relies on a sufficiently large number of labeled or supervised examples. However, in m...
Bo Wang, Jie Tang, Wei Fan, Songcan Chen, Zi Yang,...
In this paper, we define and study a new opinionated text data analysis problem called Latent Aspect Rating Analysis (LARA), which aims at analyzing opinions expressed about an e...
Probabilistic Latent Semantic Analysis (PLSA) has become a popular topic model for image clustering. However, the traditional PLSA method considers each image (document) independen...
There are many problems requiring a semantic account of a database schema. At its best, such an account consists of mapping formulas between the schema and a formal conceptual mode...