Referred Publications

Amita Misra, Brian Ecker and Marilyn Walker.
“Measuring the Similarity of Sentential Arguments in Dialogue”.
Proc. of the SIGDIAL Conference, 2016.

Amita Misra, Brian Ecker, Theodore Handleman, Nicolas Hahn and Marilyn Walker.
“NLDS-UCSC at SemEval-2016 Task 6: A Semi-Supervised Approach to Detecting Stance in Tweets”.
Proc. of the International Workshop on Semantic Evaluation, 2016.

Amita Misra, Pranav Anand, Jean E. Fox Tree and Marilyn Walker.
“Using Summarization to Discover Argument Facets in Online Ideological Dialog”.
Proc. of the NAACL-HLT2015 Conference, 2015.

Amita Misra and Marilyn Walker.
Topic independent identification of agreement and disagreement in social media dialogue.
Proc. of the SIGDIAL conference, 2013.

Projects

Implemented "Sprinkling: Supervised Latent Semantic Indexing." (2013)
as described in Advances in Information Retrieval, 28th European Conference on IR Research, ECIR 2006, London, UK, April 10-12, 2006, Proceedings, for supervised agreement disagreement classification problem.
The LSI was implemented using “R” and classification was through weka Toolkit.
Implemented for a course project.

Classification of movie scenes into different genre(2013)
The motivation was to build a movie scene database categorised into different genres such as Romance, Action and Comedy using movies titles classified by genre.
The work involved scraping the movie script website(IMSDb) using python, Part of speech tagging using NLTK toolkit, feature selection and classification using weka Toolkit.
Implemented for a course project.

Mobile App for the politicos.org.br website(2014)
The project consisted of designing a mobile App for the politicos.org.br website, which ranks Brazilian politicians based on, for example, whether they are involved in scandals and corruption charges.
The interface was built as a mock mobile App for iPhone using phonegap. It was focussed towards learning various stages in Human Communication Interface design such as persona creation, Storyboard, wireframes, competitive analysis, determining functional and nonfunctional requirement with development of low and High Fidelity prototypes.
We worked as a team of 5 members for a course project.
Language :The webcrawler was developed in python and frontend in JavaScript.