미국 내 한식당 온라인 리뷰의 감성분석을 위한 특징벡터와 분류기법의 비교 연구
장한솔, 유성준, 한동일 et al. · 대한전자공학회 학술대회 · 2013
In this paper, we did sentiment analysis for online review of Korean restaurant in Yelp using machine learning methods. We adapted various classfication algorithms to compare the performance. Also, we changed type and size of feature vectors to compare the performance. We used Adaboost, ANN, Decision Tree, kNN, Naive Bayes, SVM as classification algorithm and extracted feature vectors with PCFG Parser and POS Tagger. On average, SVM with POS Tagger showed better performance.
Cited by 0 works.