[Back]

“Research Scholar, IITB-Monash Research Academy"(January 2016 - Present)
Research Work
 - Research on Cognate Identification, under the purview of Computational Phylogenetics.
- Special focus on Indian Languages.
- Work with Sandeep Mathias on Cognitive NLP for Educational Purposes.
- Creation of Textual History Tool for digitizing manuscripts, and performing phylogenetics .

“Research Scholar, IITB-Monash Research Academy"(January 2016 - Present)
Teaching Assistant
 - Software Architecture (ongoing) with Prof. R. K. Joshi (CS718).
- Data Structures and Algorithms with Prof. R. K. Shyamasundar (CS213).
- Software Systems Lab with Prof. Kavi Arya (CS251) - Awarded TA of the Semester.
- Digital Logic Design (CS226).

“Research Scholar, IITB-Monash Research Academy"(January 2016 - Present)
Course Work (Overall CPI: 8.1)
 - Literature Survey with Seminar on Computational Phylogenetics, current tools and resources available.
- Course on Linear Optimization (CS750)
- Course on Kernel Programming (CS746)
- Course on Information and Communication Technology for Socio-Economic Development (CS671)
- Course on Artificial Intelligence (CS621)
- Course on Embedded Systems (CS684)
- Course as Software Lab (CS699)
- Course on Automatic Speech Recognition (CS753)
- Course on Software Architecture (CS718)

“Research Engineer, CFILT, IIT Bombay”(July 2013 – December 2015)
Work Profile as Research Engineer
 - Sarcasm Detection using Cognitive Studies.
- Sarcasm Understandability using Cogninitive Studies in NLP.
- Multilingual Topic Modelling to improve Indian Language Machine Translation.
- Working simultaneously on Hindi, Marathi, Sanskrit WordNets, IndoWordNet and Development of DiscrimiNET.
- DiscrimiNET - a resource based disambiguation engine with the ability to disambiguate b/w word senses based on clue words. (on a pause)
- TransChat - Machine Translation for Instant Messaging.
- Parallel Corpora Management Workbench (PaCMan) - a machine translation aid allowing translation / validation / alignment for parallel corpora.
- PanchaBhoota - Hierarchical Phrase based Statistical Machine Translation system for five language pairs.
- Cognitive studies for Sentiment annotation.
- Systems Administrator for Centre for Indian Language Technology handling / troubleshooting / maintaining both server systems, and desktop computers.

“Mechanisms for Sense Discrimination”(Jan 2013 – Mar 2013)
Improvements in “Clue Marker Tool” and its Future Implications Improved the context word searching scenario by providing the lexicographers with a rich corpus. Facilitated the clue addition to database by including Google Hindi Transliteration API within the tool, and by including Concordancer Corpus search. Automated the clue searching by providing with an automatically generated possible clue set. Work on automatic clue addition in future using PMI and DICE indexing of the clues. Laid groundwork for a possible future rule based framework to be called, “Discrimination Net”.

“Mechanisms for Sense Discrimination”(Jun 2012 – Aug 2012)
Development of “Clue Marker Tool” Emphasized on importance of context and clue words around a target word synset word. Developed “Sense Discrimination Tool” which facilitates the addition of clue words by lexicographers manually into the WordNet database. Laid groundwork for automation of the tool in future using the clues present in the database, using various heuristics.

“Man Vs. Machine Annotation for WSD”(Jan 2012 – Mar 2012)
Study of Man / Machine Annotation Process Our work aimed at understanding the meaning and significance of “contextual evidence”, with respect to the human sense disambiguation task. The experiments probed into the disambiguation time across different POS categories and ontological categories, they also investigated into the set of clues in the neighbouring words of a pre-defined target word.

"Wordnet & its applications"(July 2011 – Aug 2011)
Wordnet WordNet is a rich repository of knowledge of words and plays a key role in many Information Retrieval and Language Processing applications, includes extensive study of WordNet (English and Hindi) and various state of the art algorithms.
IWSD Algorithm Work based on Error Analysis of the IWSD algorithm, divided in various parts, viz. POS based, Ontology based, Ablation test on IWSD parameters, aim being the study of importance of contextual part in the scoring function of IWSD algorithm.