CS460/IT632 - Natural Language Processing/Language Technology for the Web

Course Instructor

Teaching Assistants

• Sagar Ranadive (sagar [AT] it.iitb.ac.in)

Time Table And Venue

Announcements

• Click here for course content and references.
• Audit requirements- To be decided
• 10-01-06: Mailing list for the course is cs460_it632@cse.iitb.ac.in. This will be used to make all the announcements etc. To subscribe to the mailing list, add your email id in the interface at this link
• 10-01-06: New link added to resources section - "Transformation Based Error Driven Learning & NLP: A case study in PoS Tagging" and Brill's Tagger
• 17-01-06: Next week lectures are on PoS tools and Work on PoS tagging for Indian languages at IITB.
• 19-01-06: New link added to resources section - Natutal Language Toolkit(NLTK_lite)
• 19-01-06: No Lecture on 20-01-2006
• 25-01-06: TnT paper covered in the last class added to the resources section.
• 13-02-06: Papers related to MEM & CRF added to the resources section.
• 24-03-06: Miniproject presentations (Stage 2) on 25-03-2006 in A2.
• 07-04-06: ENDSEM Exam: 15/04/06, 4-6 PM (2 hrs only)

Lecture Notes

• 03-01-06 - Introduction to NLP [ppt] [pdf]
• 06-01-06 - Part of Speech (PoS) Tagging [ppt] [pdf]
• 10-01-06 - Statistical Formulation of PoS Tagging Problem [ppt] [pdf]
• 13-01-06 - An Introduction to Natural Language Syntax [ppt] [pdf]
• 17-01-06 - Classical PoS Tagging [ppt] [pdf]
• 20-01-06 - No class
• 24-01-06 - Discussion of TnT paper (linked in resources section) and the current work going on in Hindi POS Tagging at IITB
• 27-01-06 - Discussion of the past work done in Hindi [pdf] and Marathi [pdf] POS Tagging at IITB
• 31-01-06 - Stemming [ppt] [pdf]
• 03-02-06 - Dealing With Corpora [ppt] [pdf]
• 07-02-06 and 10-02-06 - Graphical Models for part-of-speech tagging [ppt] [pdf]
• 14-02-06 - Top-down and Bottom-up Parsing [ppt] [pdf]
• 17-02-06 - Top-down Bottom-up Chart Parsing [ppt] [pdf]
• 03-03-06 - Prolog [ppt] [pdf]
• 04-03-06 - Language Modeling for Information Retrival (by Mr. Manoj) - [pdf]
• 07-03-06 - Language Modeling for Information Retrival (by Mr. Manoj) - [pdf]
• 10-03-06 - Miniproject presentations - Stage 1
• 14-03-06 - Formulation of Grammar And Parsing [ppt] [pdf]
• 17-03-06 - Lexical Knowledge Structures (by Mr. Ramanand) [ppt] [pdf]
• 21-03-06 - Verb Knowledge Base [pdf]
• 25-03-06 - Miniproject presentations - Stage 2
• 28-03-06 - Probabilistic Parsing [ppt] [pdf]
• 31-03-06 - Guest Lecture on Machine Translation [ppt] [pdf]
• 04-04-06 - Panel Discussion at MSPIL 2006.
• 07-04-06 - Cross Language Information Retrival (by Mr. Anand) [ppt] [pdf]
• 14-04-06 - Word Sense Disambiguation [ppt] [pdf]

Seminar & Project

• Groups Information

Assignment

• Try your hand on the NLTK_LITE which is an NLP toolkit. This has various taggers including Brill's and Statistical. You need to know Python for this, which is similar to but more elegant than Perl. Compare Brill's and statistical tagger performance for Hindi and English language. The Hindi annotated corpora is linked in the resources section.
• ASSIGNMENT 1
• ASSIGNMENT 2

Some Useful Links and Resources For Course

• Tool : Natural Language Toolkit(NLTK_lite)
• Paper : Transformation Based Error Driven Learning & NLP: A case study in POS Tagging
• Tool : Brill's Tagger
• Paper : TnT - A Statistical Part-of-Speech Tagger
• Paper : The CN2 Induction Algorithm
• CoNLL 2003 DataSet : CoNLL 2003 DataSet, The details can be found here
• CoNLL 2000 DataSet : CoNLL 2000 DataSet, The details can be found here
• Hindi POS Tagged corpora : Annotated Hindi Corpora, rest at the internal link. The TAGSET can be found here
• Paper - CRF - Lafferty : CRF : Probabilistic models for segmenting and labelling sequence data.
• Paper - CRF - McCallum : Efficiently inducing features of CRF.
• Paper - CRF - Sha & Pereira : Shallow parsing with CRF.
• Paper - MEM - Ratnaparkhi : A simple introduction to MEM for NLP.
• Paper - MEM - Ratnaparkhi : A MEM for POS Tagging.
• Paper - MEM - Pietra : A MEM approach to NLP.