Resource Construction and Evaluation for Indirect Opinion Mining of Drug Reviews

Abstract

Opinion mining is a well-known problem in natural language processing that has attracted increasing attention in recent years. Existing approaches are mainly limited to the identification of direct opinions and are mostly dedicated to explicit opinions. However, in some domains such as medical, the opinions about an entity are not usually expressed by opinion words directly, but they are expressed indirectly by describing the effect of that entity on other ones. Therefore, ignoring indirect opinions can lead to the loss of valuable information and noticeable decline in overall accuracy of opinion mining systems. In this paper, we first introduce the task of indirect opinion mining. Then, we present a novel approach to construct a knowledge base of indirect opinions, called OpinionKB, which aims to be a resource for automatically classifying people’s opinions about drugs. Using our approach, we have extracted 896 quadruples of indirect opinions at a precision of 88.08 percent. Furthermore, experiments on drug reviews demonstrate that our approach can achieve 85.25 percent precision in polarity detection task, and outperforms the state-of-the-art opinion mining methods. We also build a corpus of indirect opinions about drugs, which can be used as a basis for supervised indirect opinion mining. The proposed approach for corpus construction achieves the precision of 88.42 percent.

A Context-Based Disambiguation Model for Sentiment Concepts Using a Bag-of-Concepts Approach

“…Studies using the background knowledge in the sentiment analysis are divided into two broad categories: Studies based on domain-specific knowledge (e.g., [16], [17]), and studies dealing with commonsense knowledge(e.g., [14], [15]). Some studies have adopted ontology to create knowledge from corpora semiautomatically [18] or manually [16], [17];…”

Patient opinion mining to analyze drugs satisfaction using supervised learning

“…In recent years, we witnessed the advance in neural network methodology, like fast training algorithm for deep multilayer neural networks (Chen, Liu, & Chiu, 2011;Ghiassi, Skinner, & Zimbra, 2013;Jian et al, 2010;Luong, Socher, & Manning, 2013;Moraes, Valiati, & Neto, 2013;Noferesti & Shamsfard, 2015;Sharma & Dey, 2012). Neural network techniques have found success in several NLP tasks recently such as opinion mining (Bobicev, Sokolova, Jafer, & Schramm, 2012;Socher, Lin, Manning, & Ng, 2011).…”

Automatic lexicon construction for domain-specific sentiment analysis : a frame-based approach

“…One of the many difficulties faced by the sentiment analysis research community is that sentiment can be expressed in many different ways, some of which involve the use of opinionated terms that convey attitude in a relatively straightforward manner while others contain no prominent sign of evaluative or emotion-related words but still express sentiment implicitly (Balahur, Hermida, Montoyo, & Muñoz, 2013;Cambria, Schuller, Xia, & Havasi, 2013;Liu, 2015;Noferesti & Shamsfard, 2015;Ofek et al, 2016). Dealing with implicit sentiment expression has been a lingering concern in sentiment analysis research.…”

“…Unlike earlier research in lexicon construction that focused primarily on strongly evaluative or emotion-related terms like "good," "bad," "awful," "excellent," and so forth (Blair-Goldensohn et al, 2008;Hu & Liu, 2004;Kamps et al, 2004;Rao & Ravichandran, 2009;Wilson et al, 2005), this research emphasizes the importance of phrasal lexicon entries, which play a vital role in detecting implicit sentiments and indirect opinions (Balahur et al, 2013;Cambria et al, 2012Cambria et al, , 2014Cambria et al, , 2016Cambria et al, , 2018Noferesti & Shamsfard, 2015;Ofek et al, 2016). By addressing the difficulties in extracting phrasal lexicon entries from text corpora, this research provides an important opportunity to advance our understanding of the complex problem of implicit sentiment analysis.…”

“…The idea of extracting phrasal patterns at the semantic-level offers some new insights into how domain lexicon construction can benefit from the semantic similarity that transcends lexical-level variation. Since phrasal units play a crucial role in detecting implicit sentiments and indirect opinions (Balahur et al, 2013;Cambria et al, 2012Cambria et al, , 2014Cambria et al, , 2016Cambria et al, , 2018Noferesti & Shamsfard, 2015;Ofek et al, 2016), new understanding and avenue in dealing with the extraction of phrasal units should help to foster progress on this complex problem of sentiment analysis.…”