2013 Annual Meeting
Montrťal, Quťbec, Canada | November 1-5, 2013
Shuqing Li, Nanjing University
Ying Sun, University of Buffalo
The discovery and visualization of temporal sequence of personalized academic research can enhance the ability for discovering the latent trend of usersí interests. In this paper, we propose a definition of weighted co-occurred keywords time gram and use it as a basic unit to analyze the temporal information in existed keywords collection. We further propose a method to get the temporal sequence and temporal network based on these time grams. An application of the proposed method in discovering academic research temporal sequence is discussed, which includes techniques for acquiring extended keywords, assigning weight to each keyword and co-occurred weight to each keyword pair. A visualization tool is designed for browsing the temporal networks identified. Finally, we report an experiment in the area of library and information studies. The experiment results show the effectiveness of the proposal method in helping users analyzing and portraying the evolution pattern and developing trend of corresponding academic research.