IIIT logo Prof. Rajeev Sangal
Professor
International Institute of Information Technology - Hyderabad
India

1.Overview
2.Teaching
3.Research Interests
        3.1 Natural Language Processing
        3.2 Artificial Intelligence
        3.3 New Curricula in Computer Science
        3.4 Humanities
        3.5 Challenges in Higher Education
        3.6 Books
4.Professional Activities
5.Links Worth Visiting

3.1X Natural Language Processing:
        3.1.1 Computational Grammatical Models
        3.1.2 Natural Language Parsing
        3.1.3 Machine Translation
        3.1.4 Discourse & Dialogue Systems
        3.1.5 Other NLP Areas
3.2X Artificial Intelligence


1. Overview

I work on processing of human languages in Computational Paninian Grammar (CPG) framework, particularly applied to Indian languages. Our team is building machine translation systems among Indian languages and from English to Indian languages. I also work on knowledge representation, information extraction, search engines. I have authored 3 books, and a number of research papers in NLP and AI.

I have been fortunate to be able to play a major role in shaping a new research university, namely IIIT Hyderabad. It has research centres but not departments. It focuses on research, while at the same time looking to applicability of technologies developed. It promotes research through its unique environment of experimentation and openness. I have developed and successfully implemented a radical new curriculum in Computer Science and Engg., that promotes research and project work starting from an early stage at the undergraduate level.

I have helped make human values course a regular part of the academic curriculum at IIIT-H, where it is playing a major role in shaping the atmosphere, and relating to larger human and societal concerns.

(View more...)

TOP



2. Teaching

TOP



3. Research Interests

    3.1 Natural Language Processing

    I (along with my senior colleague Dr Vineet Chaitanya) have developed Computational Paninian Grammer (CPG) for Indian languages and beyond. It is a dependency model based on Paninian insights applied to modern Indian languages.

    I also work on parsing of free word-order languages, including Indian languages. Work includes parsing using Computational Paninian Grammar (CPG) framework and machine learning from treebank.

    I lead a consortium of 12 academic institutions to build Machine Translation Systems for 9 Indian language pairs. These are bi-directional and hence 18 systems in all.

    I work on discourse & dialogue, specifically on Natural Language Interface to Databases (NLIDB systems). The aim is to develop a portable system, across languages and domains, so that people without linguistic expertise and basic technical knowledge can use such a NLP application.

    For details of the different subareas, see below in 3.1X.

    3.2 Artificial Intelligence

    I am interested in Knowledge Representation and Ontology particularly on identifying the right level of both which helps one connect with NLP. PurposeNet is an outcome of that research. See below in 3.2X.

    3.3 New Curricula in Computer Science
    3.4 Humanities
    3.5 Challenges in Higher Education
    Some of my experiences in building a university are given in presentations
    1. Building a Research University: A Case Study of IIIT Hyderabad
    2. Tenth Anniversary Presentation (link to be added)

    3.6 Books
    I have also authored books in the field computer science and humanities.
    1. A Foundation Course in Human Values and Professional Ethics
      RR Gaur, Rajeev Sangal, GP Bagaria
      Excel Books, New Delhi, 2010.
    2. Natural Language Processing: A Paninian Perspective
      Akshar Bharati, Vineet Chaitanya and Rajeev Sangal
      Prentice-Hall of India, New Delhi, 1995.
    3. LISP Programming
      Rajeev Sangal
      Tata-McGraw Hill, New Delhi, 1995.
    4. Programming Paradigm in LISP
      Rajeev Sangal
      McGraw Hill, New York, 1991. (International student edition by McGraw Hill-Kogakusha, Singapore, 1992.)
TOP



4. Professional Activities

TOP



6. Links Worth Visiting

TOP



3.1X Natural Language Processing:

3.1.1 Computational Grammatical Models
I have worked on developing Computational Paninian Grammer (CPG) for Indian languages and beyond. I have done extensive work on dependency model based on Paninian insights and applied it to modern Indian languages.

Our book of 1995 summarises the research done on CPG till then at one place.
"Natural Language Processing : A Paninian Perspective" By Akshar Bharati, Vineet Chaitanya and Rajeev Sangal.
Prentice Hall of India, 1995. Download
The following papers are on CPG :
  1. External Sandhi and Its Relevance to Syntactic Treebanking
    Sudheer Kolachina, Phani Gadde, Meher Vijay, Dipti Misra Sharma, Rajeev Sangal
    In proceedings of 12th Conference on Intelligent Text Processing and Computational Linguistics, Japan, February 2011
  2. Dependency Annotation Scheme for Indian Languages
    Rafiya Begum, Samar Husain, Arun Dhwaj, Dipti Misra Sharma, Lakshmi Bai and Rajeev Sangal
    Proceedings of The Third International Joint Conference on Natural Language Processing (IJCNLP). Hyderabad, India. 2008
          (View more papers...)

3.1.2 Natural Language Parsing
I also work on parsing of free word-order languages, such as Indian languages.
Work includes parsing using Computational Paninian Grammar (CPG) framework and machine learning from treebank.
  1. Clausal Parsing helps data-driven dependency parsing: Experiments with Hindi
    Husain Samar, Phani Gadde, Joakim Nivre, Rajeev Sangal
    In proceedings of 5th International Joint Conference on NLP, Thailand (IJCNLP), 2011
  2. On the Role of Morphosyntactic Features in Hindi Dependency Parsing
    Bharat Ambati,Samar Husain,Joakim Nivre,Rajeev Sangal
    NAACL 2010: Human Language Technologies: The 11th Annual Conference of the North American Chapter of the Association for Computational Linguistics
  3. Two methods to incorporate local morphosyntactic features in Hindi dependency parsing
    Bharat Ambati,Samar Husain,Sambhav Jain,Dipti Misra Sharma,Rajeev Sangal
    NAACL 2010: Human Language Technologies: The 11th Annual Conference of the North American Chapter of the Association for Computational Linguistics
  4. Improving Data Driven Dependency Parsing using Clausal Information
    Phani Gadde,Karan Jindal,Samar Husain,Dipti Misra Sharma,Rajeev Sangal
    11th Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL-HLT, 2010)
          (View more papers...)


3.1.3 Machine Translation
  1. Consortium Leadership :
    I lead a consortium of 9 academic institutions to build Machine Translation Systems for 9 Indian language pairs. These are bi-directional and hence 18 systems in all.
    To try out the released systems click on : Sampark
    1. Re-engineering Machine Translation Systems through Symbiotic Approach
      Pawan Kumar,Rashid Ahmad,Arun Kumar,Mukul K Sinha,Rajeev Sangal
      In proceedings of 3rd International Conference of Contemporary Computing (IC3 2010)
    2. Handling Multi-word Expressions without Explicit Linguistic Rules in an MT System
      Akshar Bharati, Rajeev Sangal, Dipti M Sharma, Sriram V and T Papi Reddy
      Proceedings of Seventh International Conference on TEXT, SPEECH and DIALOGUE - 2004, Brno, Czeck Republic
              (View more papers...)

  2. I have also worked to develop Shakti Machine Translation from English to Hindi
  3. Earlier I have also used Anusaaraka Models to build language access systems across Indian languages during 1991 - 1998

  4.    The following are research papers :
    1. Anusaaraka: Overcoming the Language Barrier in India
      Akshar Bharati, Amba P Kulkarni, Vineet Chaitanya, Rajeev Sangal and G. Umamaheshwara Rao
              (View more papers...)


3.1.4 Discourse & Dialogue Systems
I am working in this area, specifically on Natural Language Interface to Databases (NLIDB systems). The aim is to develop a portable system, across languages and domains, so that people without linguistic expertise and basic technical knowledge can use such a NLP application.
  1. A Novel Approach Towards Building a Portable NLIDB System Using the Computational Paninian Grammar Framework
    Abhijeet Gupta, Arjun Akula, Deepak Malladi, Puneeth Kukkadapu, Vinay Ainavolu, Rajeev Sangal
    Published in "IALP-2012: International Conference on Asian Language Processing, Hanoi - Vietnam"
  2. A Novel Approach for Tagging of Discourse Segments in Help-desk Emails
    Suman Yelati, Rajeev Sangal
    Next Trends in Email Workshop of WI-IAT, France, September 2011

3.1.5 Other NLP Areas

  1. Shallow Parsing of Indian Languages
    1. HMM Based Chunker for Hindi
      Akshay Singh, S M Bendre and Rajeev Sangal
      Proceedings of IJCNLP-05: The Second International Joint Conference on Natural Language Processing, 11-13 October, 2005, Jeju Island, Republic of Korea
              (View more papers...)

  2. Unsupervised Parsing
    1. Two Approaches for Building an Unsupervised Dependency Parser and Their Other Applications
      Jagadeesh Gorla, Amit Goyal, Rajeev Sangal.
      AAAI 2007. pp.1860-1861.
              (View more papers...)

  3. Corpus Processing
    1. Basic Statistical Analaysis of Corpus and Cross Comparision
      Akshar Bharati, Prakash Rao, Rajeev Sangal and S M Bendre
      Proceedings of ICON-2002: International Conference on Natural Language Processing, Mumbai, 18-21 Dec 2002
              (View more papers...)

  4. Font conversion and learning
    1.           (View more papers...)

    2. Others
      1. Dashboard: An Integration and Testing Platform based on Blackboard Architecture for NLP Applications
        Pawan Kumar,Arun Kumar,Rashid Ahmad,Mukul K Sinha,Rajeev Sangal
        6th International Conference on Natural Language Processing and Knowledge Engineering (IEEE NLP-KE 2010)
      2. Unit Selection Voice for Amharic Using Festvox
        Sebsibe H/Mariam, Kishore S.P., Rohit Kumar, Alan W Black and Rajeev Sangal
        Published in the Proceedings of 5th ISCA Speech Synthesis Workshop,14-16 June 2004, Pittsburgh, PA
                (View more papers...)

    3.2X Artificial Intelligence

    1. Hindi Semantic Category Labeling Using Semantic Relatedness Measures
      Siva Reddy,Abhilash Inumella,Navjyoti Singh,Rajeev Sangal
      Global Wordnet Conference 2010 (GWN2010 2010)
    2. An Ontological Resource Organized Around Purpose
      P Kiran Mayee, Rajeev Sangal, Soma Paul and Navjyoti Singh
      Published in "ICON-2008: 6th International Conference on Natural Language Processing, Pune 2008
    3. A New Inference Procedure for Conceptual Graphs
      Akshar Bharati, S.A. Nawathe, Vineet Chaitanya and Rajeev Sangal
      4th University of New Brunswick Artificial Intelligence Symposium, 1991d, 19-21 Sept., Frederiction, N.B. Canada

    TOP


    Last Updated: December, 2012
    Page Format Updated: March, 2012