Abstract
This paper reports our submissions to the four subtasks of Aspect Based Sentiment Analysis (ABSA) task (i.e., task 4) in SemEval 2014 including aspect term extraction and aspect sentiment polarity classification (Aspect-level tasks), aspect category detection and aspect category sentiment polarity classification (Categorylevel tasks). For aspect term extraction, we present three methods, i.e., noun phrase (NP) extraction, Named Entity Recognition (NER) and a combination of NP and NER method. For aspect sentiment classification, we extracted several features, i.e., topic features, sentiment lexicon features, and adopted a Maximum Entropy classifier. Our submissions rank above average.
Highlights
Sentiment analysis has attracted a lot of attention from researchers
We found that the noun phrase (NP)-based method generates many noisy terms resulting in high recall and low precision, and the Named Entity Recognition (NER)-based method performs inverse results
Our system ranks above the average under constrained model, which proves the effectiveness of the combination method by using NP extraction and NER
Summary
Most previous work attempted to detect overall sentiment polarity on a text span, such as document, paragraph and sentence. Since sentiments expressed in text always adhere to objects, it is much meaningful to identify the sentiment target and its orientation, which helps user gain precise sentiment insights on specific sentiment target. The aspect based sentiment analysis (ABSA) task (Task 4) (Pontiki et al, 2014) in SemEval 2014 is to extract aspect terms, determine its semantic category, and to detect the sentiment orientation of the extracted aspect terms and its category. The aspect term extraction (ATE) aims to extract the aspect terms from the sentences in two giv-
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
