In recent years, South-Asian literature in English has experienced a surge of newfound love and popularity both in the local and the global market. In this regard, Arundhati Roy's The God of Small Things (1997) has garnered an astounding mix of positive and negative reactions from readers across the globe. This chapter adopts an artificial intelligence approach to analyse netizen readers' feedback on the novel as documented in the book cataloguing website Goodreads. To this end, an opinion mining framework is proposed based on artificial intelligence techniques such as topic modelling and sentiment analysis. Latent semantic analysis (LSA) and latent Dirichlet allocation (LDA) are applied and compared to find the abstract “topics” that occur in a collection of reviews. Furthermore, lexicon-based sentiment analysis approaches such as Vader and Textblob algorithms are used and compared to find the review sentiment polarities.
TopIntroduction
South-Asian Literature in English (SLE hereafter) refers to a broad collection of literary works written in English encompassing the people and the topics of South-Asian countries such as Pakistan, Bangladesh, Sri Lanka, and India. Over the past few decades, there has been a surge of newfound love and popularity for SLE both in the local and the global market (Asl, 2022; Dwivedi & Lau, 2014). The recent development of SLE is a result of “a new kind of social experience, born out of South Asia’s accelerated economic and demographic growth, its global reach and its complex internal and regional politics” that has permeated the lives of many, and is demanding fictional representation suited to represent these changes (Tickell, 2016, p. 2). In other words, SLE is seen as a colossal body of literary works continuously reforming and adapting to reflect present issues that shape reality and social fabrics of South-Asian communities. ‘Colossal’, in particular, is used to describe SLE for its inclusion of literary works not only by home-based writers but also by diaspora writers.
Among the woman writers of SLE, Arundhati Roy is a name widely known for being a prolific home-based Indian woman writer writing about gender issues and politics. She has published two novels relating to these subjects—The God of Small Things in 1997 and The Ministry of Utmost Happiness in 2017. The God of Small Things (1997)—TGST hereafter—tells a story about family and social obligation, Indian politics and society and class, and love and sexuality. Since its publication, TGST has garnered attention from critics and readers alike. In 1997, it has won the Booker Prize award. Expectedly, the reception of Roy’s debut novel is an astounding mix of positive and negative reactions. Nevertheless, it has been noted by critics that the responses are so extreme and polarized that it is difficult to believe that they are reading the same work (Anuar & Asl, 2021). Tickell (2007) provides that the two opposing groups would either love the novel deeply or dislike it with passion:
admiring readers describe [an] almost mystical attachment to her fiction and regard the novel as ‘magical’, ‘breathtakingly beautiful’ and ‘close to perfection’. Many also note the book’s emotional impact and its lingering ‘imprint’ on the reader, and others talk perceptively about the fantastic, interlocking musical patterns of Roy’s writing, its descriptive originality and the way key words and phrases evoke specific moods and events. In the opposite camp, Roy’s less appreciative readers repeatedly attack the novel’s unwarranted ‘hype’, its ‘tediously’ overwritten or needlessly embellished style and the difficulty of following the plot through its fragmented time scheme. (p. xiii)
In other words, the audience of TGST are so divided in their opinions regarding the work that the same elements which make the novel ‘beautiful’ are also seen as ‘tedious’ by different critics and readers.
With the advancement of technology and the emergence of book cataloguing websites such as Goodreads, literary readers and enthusiasts across the world have found a unique platform to share their reviews of literary works. For instance, when one opens the homepage of the book in Goodreads, one can see 16,904 reviews and 276,016 ratings (as of 15 November 2022) documented for the book. Apparently, the amount of data in such websites created every day is enormous since internet-based platforms now allow millions of people to exchange information. The issue is how this enormous amount of data that includes conflicting views can be processed. Here is where text mining is revealed which aims to extract useful information for a specific purpose by analyzing the text (Keikhosrokiani & Asl, 2022; Suhendra et al., 2022). This type of analysis is often carried out manually by literary critics and readers which poses certain methodological difficulties (Chu et al., 2022; Malik et al., 2021; Yun Ying et al., 2022).