A Study of Suggestions in Opinionated Texts and their Automatic Detection

Proceedings of the Fifth Joint Conference on Lexical and Computational Semantics (*SEM 2016), pages 170–178, Berlin, Germany, August 11-12, 2016. A Study of Suggestions in Opinionated Texts and their Automatic Detection Sapna Negi1 Kartik Asooja1 Shubham Mehrotra1,2 Paul Buitelaar1 1 Insight Centre for Data Analytics, National University of Ireland, Galway firstname.lastname@insight-centre.org 2 Indian Institute of Information Technology Allahabad, India shubhammehrotra94@gmail.com Abstract We study the automatic detection of sug- gestion expressing text among the opin- ionated text. The examples of such sug- gestions in online reviews would be, cus- tomer suggestions about improvement in a commercial entity, and advice to the fel- low customers. We present a qualitative and quantitative analysis of suggestions present in the text samples obtained from social media platforms. Suggestion min- ing from social media is an emerging re- search area, and thus problem definition and datasets are still evolving; this work also contributes towards the same. The problem has been formulated as a sentence classification task, and we compare the re- sults of some popular supervised learning approaches in this direction. We also eval- uate different kinds of features with these classifiers. The experiments indicate that deep learning based approaches tend to be promising for this task. 1 Introduction Online text is becoming an increasingly popu- lar source for acquiring public opinions towards entities like persons, products, services, brands, events, etc. The area of opinion mining focuses on exploiting this abundance of opinions, by mainly performing sentiment based summarisation of text into positive, negative, and neutral categories, us- ing sentiment analysis methods. In addition to the online reviews and blogs, people are increasingly resorting to social networks like Twitter, Facebook etc. to instantly express their sentiments and opin- ions about the products and services they might be experiencing at a given time. On a closer look, it is noticeable that opinion- ated text also contains information other than sen- timents. This can be validated from the presence of large portions of neutral or objective or non- relevant labelled text in state of the art sentiment analysis datasets. One such information type is suggestions. Table 1 shows the instances of sug- gestions in sentiment analysis datasets which were built on online reviews. These suggestions may or may not carry positive or negative sentiments to- wards the reviewed entity. In the recent past, sug- gestions have gained the attention of the research community, mainly for industrial research, which led to the studies focussing on suggestion detec- tion in reviews (Ramanand et al., 2010; Brun and Hagege, 2013). The setting up of dedicated suggestion collec- tion forums by brand owners, shows the impor- tance of suggestions for the stakeholders. There- fore, it would be useful if suggestions can be au- tomatically extracted from the large amount of al- ready available opinions. In the cases of certain entities where suggestion collection platforms 1 are already available and active, suggestion min- ing can be used for summarisation of posts. Often, people tend to provide the context in such posts, which gets …

Content truncated to 500 words