Output details
11 - Computer Science and Informatics
Aston University
Joint sentiment/topic model for sentiment analysis
<22> The research represents a breakthrough in simultaneously detecting topics and topic-associated sentiments from text without the use of labelled data. Since the proposal of the joint sentiment-topic (JST) model, there have been several papers extending JST (e.g., Jo and Oh, 2011, DOI 10.1145/1935826.1935932 and Ramage et al., 2011, DOI 10.1145/2020408.2020481).
The model has been integrated into Luxid Skilled Cartridge, a text mining product developed by TEMIS, a text analytics company with 7 branches, in an EC-FP7 project, ROBUST (grant number 257859). TEMIS is discussing commercialising JST (Stefan Geißler, Managing Director of TEMIS Germany, stefan.geissler@temis.com) with Aston.