Oct 12, 2015 | Asmita Poddar
The department of Computer Science and Engineering is always geared up to explore new fields of science and technology, and NITR reached new heights internationally as Prof. Korra Sathya Babu recently attended a conference called ASONAM ( International Conference on Advances in Social Networks Analysis and Mining) hosted by Telecom Paris. Prof. Korra Sathya Babu presented a paper on 'Detection of Sarcastic sentiments in Twitter Data' there.
When some reviews are given about a product on a social networking forum like Twitter or Facebook or any other review blogs, it is not always given in a direct way. For example, a customer might not want to explicitly attack a company by his/her review, and might give a review in a veiled manner. The customers may give a negative review using positive words, in other words, sarcastically. Due to such review the polarity of the product changes, even though the review contains positive words.
However, a company needs to assess the feedback given by the customers for predicting how well a product is doing, or where they need to focus on to make a product more successful or where they are going wrong with particular product. All this plays a major role in impacting the sales and profits of a company.
There are already systems in place, which can detect the key words like 'good', 'excellent' or 'not good', to figure out what the general review is actually conveying- a positive or a negative reaction. What we tried to do is detect some patterns using some conventional computation techniques, use those patterns to identify sarcastic comments and show that the seemingly positive reviews are actually negative.
Telecom Paris is a prestigious institute, and in fact, a scientist of that institute coined the term 'Telecom'. The conference saw many notable scholars and researchers from all over the world. Prof. Redaalhajj from the University of Calgary was the Chair at the conference. There were a number of talks from people from both industry and academia.
Prof. Korra's own area of research is data science, data mining, sentiment detection, data privacy and social computing, which includes many aspects of computational sciences in society. Sarcasm Detection on Social Networks was an integral part of social computing.