Research

My doctoral research was in the area of Natural Language Processing. The objective of my research was to use cognitive information to improve the performance of automatic essay grading systems. I made use of the Eye-Tracking technology to record a reader's gaze behaviour. As recording gaze behaviour at run time is also expensive and cumbersome, I also worked on ways to learn gaze behaviour without the need for recording it at run time. To know more, please visit the Cognitive NLP website.

Publications

Conference Papers

  1. Rahul Kumar, Sandeep Mathias, Sriparna Saha, and Pushpak Bhattacharyya. Many Hands Make Light Work: Using Essay Traits to Automatically Score Essays. NAACL 2022. Seattle, WA, USA. 10 July - 15 July, 2022. [Paper] [Data] [Code]

  2. Sandeep Mathias, Diptesh Kanojia, Abhijit Mishra, and Pushpak Bhattacharyya. A Survey on Using Gaze Behaviour for Natural Language Processing. IJCAI 2020. Kyoto, Japan (Online). 7 January - 15 January, 2021. [Paper]

  3. Sandeep Mathias, Rudra Murthy, Diptesh Kanojia, and Pushpak Bhattacharyya. Cognitively Aided Zero-Shot Automatic Essay Grading. ICON 2020. Patna, India (Online). 18 December - 21 December, 2020. [Paper][Code]

  4. Sandeep Mathias, Rudra Murthy, Diptesh Kanojia, Abhijit Mishra, and Pushpak Bhattacharyya. Happy Are Those Who Grade without Seeing: A Multi-Task Learning Approach to Grade Essays Using Gaze Behaviour. AACL-IJCNLP 2020. Suzhou, China (Online). 4 December - 7 December, 2020. [Paper][Code][Data]

  5. Sandeep Mathias, Diptesh Kanojia, Kevin Patel, Samarth Agrawal, Abhijit Mishra, and Pushpak Bhattacharyya. Eyes are the Windows to the Soul: Predicting the Rating of Text Quality Using Gaze Behaviour. ACL 2018. Melbourne, Australia. 15 July - 20 July, 2018. [Paper][Data]

  6. Sandeep Mathias, and Pushpak Bhattacharyya. ASAP++: Enriching the ASAP Automated Essay Grading Dataset with Essay Attribute Scores. LREC 2018. Miyazaki, Japan. 7 May - 12 May, 2018. [Paper][Data]

Workshop Papers

  1. Sandeep Mathias, and Pushpak Bhattacharyya. Can Neural Networks Automatically Score Essay Traits?. NLP-BEA 2020. Seattle, USA. 10 July, 2020. [Paper]

  2. Sandeep Mathias, and Pushpak Bhattacharyya. Thank "Goodness"! A Way to Measure Style in Student Essays. NLP-TEA 2018. Melbourne, Australia. 19 July, 2018. [Paper]

  3. Nikhil Wani, Sandeep Mathias, Jayashree Aanand Gajjam, and Pushpak Bhattacharyya. The Whole is Greater than the Sum of its Parts: Towards the Effectiveness of Voting Ensemble Classifiers for Complex Word Identification. NLP-BEA 2018. Seattle, USA. 5 June, 2018. [Paper]

  4. Sandeep Mathias, and Pushpak Bhattacharyya. How Hard Can it Be? The E-Score - A Scoring Metric to Assess the Complexity of Text. QATS 2016. Portoroz, Slovenia. 28 May, 2016. [Paper]

  5. Sandeep Mathias, and Pushpak Bhattacharyya. Using Machine Translation Evaluation Techniques to Evaluate Text Simplification Systems. QATS 2016. Portoroz, Slovenia. 28 May, 2016. [Paper]

Projects

Significant

  1. Cognitively Aided Automatic Essay Grading

    (Duration: July, 2014 - July, 2020, Status: Completed (Ph.D. Thesis))

    The research in this thesis aims to develop a system which can automatically score essays, written by students (called an automati essay grading (AEG) system). In this thesis, my guide and I explored ways to use cognitive information to aid AEG systems. This cognitive information is extracted through the gaze behaviour of the reader.

    One of the first pieces of our work was to show a proof-of-concept that gaze behaviour can help a system assess the quality of the text. This work showed that using gaze behaviour can help a system improve its prediction of how a reader would rate the quality of the text.

    The aim of AEG systems is to score essays without having a reader read the text. Therefore, one of the main challenges was “How do we record the gaze behaviour of a reader without the reader having to read the text?” There are different ways in which this can be accomplished. Some researchers used type aggregates for each token from an existing gaze behaviour dataset. We used multi-task learning to learn the gaze behaviour of readers, using a small amount of seed data.

    Another aspect of AEG which we looked at was trait-specific automatic essay grading, where, instead of scoring the entire essay, we score individual essay traits, like content, organization, style, word choice, sentence fluency, etc. This is important, as it can provide useful feedback to the essay's writer about their work. Part of this work was done jointly with a student of IIT Patna, named Rahul Kumar.

  2. Text Simplification

    (Duration: July, 2014 - May, 2016, Status: Completed)

    The aim of this research was to find ways to quantify the complexity of text and find ways to simplify the text. Some of the techniques that I explored was sentence splitting, lexical substitution, etc.

Quizzes

One of my hobbies is quizzing. During my time at IIT Bombay, I have conducted a number of quizzes. Here is a list of them. You can go through them at your leisure, but bear in mind that I will not be making any corrections / updates to the quizzes. This is mainly to target people who shamelessly copy my questions without doing the research that I put into them. You can copy them, but at your own peril.

  1. Cricket Quiz. Conducted in February, 2015.

Quiz Original Date
Genre
Comments
Mood Indigo Lone Wolf, 2013 December, 2013 General Solo Inter-College General Quiz
Mood Indigo Lone Wolf, 2015 December, 2015 General Solo Inter-College General Quiz
IITB Solus Rex, 2013 March, 2014 General Solo Intra-College General Quiz
IITB Solus Rex, 2014 November 2014 General Solo Intra-College General Quiz
IITB Solus Rex, 2016 October 2016 General Solo Intra-College General Quiz
IITB Solus Rex, 2019 March 2019 General Solo Intra-College General Quiz
IITB Ruckus Tangdi, 2016 January 2017 General Random Team Intra-College General Quiz
IITB Ruckus Tangdi, 2017 October 2017 General Random Team Intra-College General Quiz
IITB Ruckus Tangdi, 2018 October 2018 General Random Team Intra-College General Quiz
All Things Blue September 2015 General A Quiz on Blue stuff
Ruckus Tangdi Quizzing League September 2019 - March 2020 2020 General Quizzing League of 6 quizzes
That's All Toons! March 2014 Entertainment Written quiz on cartoons, animated films, etc.
Century of Cinema Quiz January 2016 Entertainment Quiz on movies from 1915 to 2015
MELA Quiz January 2019 Entertainment Intra-College entertainment quiz
Cricket Quiz February 2015 Sports Quiz on Cricket

More quizzes will be added later on.

Dataset and Resources

1. Cognitive NLP:

Eye-tracking datasets for various NLP and Psycholinguistic tasks viz. Sentiment Analysis, Sarcasm Detection, Coreference Resolution, Text Quality Assessment, and Text Readability Assessment can be downloaded from this website (Go to “Resources”).

2. Essay Trait Scores:

Dataset for “ASAP++ Essay Trait Scores” here

Contact

  • mathiassandeep[at]gmail[dot]com
  • +91-1000010000110011011110000000011001000111 (Spambots' bliss!!!)
  • A 802, Nagarjuna Meadows, Yelahanka-Doddaballapur Road, Bangalore-560064, Karnataka, India.
  • Skype Me