Improvement of sentiment analysis via re-evaluation of objective words in SenticNet for hotel reviews

  • PDF / 301,592 Bytes
  • 11 Pages / 439.37 x 666.142 pts Page_size
  • 29 Downloads / 205 Views

DOWNLOAD

REPORT


Improvement of sentiment analysis via re-evaluation of objective words in SenticNet for hotel reviews Chihli Hung1 • Wan-Rong Wu1 • Hsien-Ming Chou1

Accepted: 13 October 2020 Ó Springer Nature B.V. 2020

Abstract In order to extract the correct sentiment polarity from word of mouth (WOM), a wide-scale and well-organized sentiment lexicon is generally beneficial. SenticNet is one such lexicon. However, it consists of a high proportion of objective words, which are generally considered to be of little use for sentiment classification due to their ambiguity and lack of sentiments. In the literature, there is a dearth of models that focus on this issue. An objective word appearing more frequently in positive sentences than in negative sentences implies a strong relationship in a positive sentiment orientation, and conversely, an objective word appearing more frequently in negative sentences implies a strong relationship in a negative sentiment orientation. Thus, the ratio of an objective word appearing in positive and negative sentences provides the sentiment orientation. Based on this concept, this paper re-assigns the sentiment values to the objective words in SenticNet and builds a revised SenticNet. Three classification techniques, the J48 decision tree, support vector machine, and multilayer perceptron neural network are used for classification. According to the experiments, the proposed models which extract sentiment values from the revised SenticNet, significantly outperform those models which extract sentiment values from the original non-revised SenticNet. Keywords Sentiment analysis  SenticNet  Sentiment lexicon  Word of mouth  Objective word & Hsien-Ming Chou [email protected] Chihli Hung [email protected] Wan-Rong Wu [email protected] 1

Department of Information Management, Chung Yuan Christian University, Taoyuan City, Taiwan

123

Lang Resources & Evaluation

1 Introduction This paper proposes a novel approach, which provides an improvement of sentiment classification through the re-evaluation of objective words in SenticNet 3 (Cambria et al. 2014) for Word of Mouth (WOM) in the hotel domain. WOM or opinionated text is a text message regarding a consumer’s experiences about a product, a service, a brand, etc., posted on the Internet. Potential consumers have become accustomed to referencing hotel WOM documents when planning a trip due to the proliferation of hotel review websites. Generally speaking, WOM is more influential than commercial advertising (Fu et al. 2015; Sa´nchez-Rada and Iglesias 2016) and is considered to be a form of crowd intelligence (Xia et al. 2016). A positive WOM encourages a consumer to make a positive decision while a negative WOM dissuades a consumer from buying. Due to the popularity of social media on the Internet, the question of how to mine opinions from WOM efficiently and effectively has attracted the attention of practitioners and researchers in the field of sentiment analysis or opinion mining. Sentiment classification based on sentiment polarities of WOM is