Seminar on Natural-Language Processing in the winter term 2011/2012

General Information

Subjects	syntax-based statistical machine translation und weighted tree languages and translations
Audiences	committed students who study Bachelor Informatik, Master Informatik, or Diplom Informatik (PO 2004 or PO 2010), and who want to take a Proseminar or Hauptseminar Master Computational Logic, and who want to take a seminar
Requirements for participation	none; basic knowledge about the subjects is recommended
Requirements for passing	on your own initiative and on the due date, make appointments with your supervisor (at least 1 week in advance) and hand in the required material (not for Proseminar) a seminar essay of 12–15 pages, complete with title, author, introduction (1 page min.), complete references, self-contained regarding notions and notations, examples and illustrations; of this essay, a preliminary version: complete regarding content, but rudimentary in presentation (Proseminar) Hand-out 1–2 pages (just one sheet!) (everybody) Talk of 30–35 minutes, supported with the use of suitable media: slides, black board, transparencies, hand-out etc.; of everything, a preliminary version: complete regarding content, but rudimentary in presentation presence at all talks, active participation at the discussions for inclusion into module examination: survey knowledge of the seminar contributions (core statements)

Schedule

Attention: meetings take place in INF 3027!

date	event
week of October 10	initial meeting and topic assignment: October 14, 13:00, INF 3027
until November 4	appointments with superviser (to be arranged in due time!); aims: being able to sketch the problem and approaches of your topic, and to answer questions in that regard; otherwise: being able to ask concrete questions; in case of major problems, make appointment before October 28!
until November 18	hand in preliminary version of your essay, make appointment for its discussion
until December 2	hand in finished essay
until December 16	preliminary version of all your presentation material, make appointment for its discussion
until January 13	hand in finished presentation material
January 25, 08:45–14:00	final meeting: talks (including a one-hour lunch break)

Active Topics

No.	Title	Literature	Supervisor	Student
1	Discriminative Training: Motivation report • slides	[1]	Büchse	Ahmed Irfan
4	Minimum-Error-Rate Training report • slides	[4]	Büchse	Wenqian Wang
8	Multidimensional Trees report • slides	[8a, 8b]	Dietze	Juliana Hildebrandt
10	Probabilistic Linear Context-free Rewriting Systems (PLCFRS) report • slides	[10a, 10b]	Dietze	Tobias Denkinger
11	Viterbi Training for PCFGs report • slides	[11]	Stüber	Franz Gregor

Literature

Some downloads only work from within the university network.

[1]	Michael Collins, 2005. Parameter Estimation for Statistical Parsing Models: Theory and Practice of Distribution-Free Methods. In New Developments in Parsing Technology. Text, Speech and Language Technology, 2005, Volume 23, Pages 19–55, Springer.	pdf
	Sections 1–4
[2]	Michael Collins, 2005. Discriminative Reranking for Natural Language Parsing. In Computational Linguistics, March 2005, Volume 31, Number 1, Pages 25–70, ACL.	pdf
	Sections 4, 5.2
[3]	Michael Collins, 2002. Discriminative Training Methods for Hidden Markov Models: Theory and Experiments with Perceptron Algorithms. In Proceedings EMNLP 2002, ACL.	pdf
	Sections 3, 5
[4]	Franz Josef Och, 2003. Minimum-Error-Rate Training in Statistical Machine Translation. In Proceedings ACL 2003, ACL.	pdf
	Sections 1–5
[5]	Libin Shen, Anoop Sakar, and Franz Josef Och, 2004. Discriminative Reranking for Machine Translation. In Proceedings HLT-NAACL 2004, ACL.	pdf
	Sections 1, 3 (up to end of 3.3), 4 (up to end of 4.1)
[6]	Percy Liang, Alexandre Bouchard-Côté, Dan Klein, and Ben Taskar, 2006. An End-to-End Discriminative Approach to Machine Translation. In Proceedings ACL 2006, ACL.	pdf
	Sections 1, 2, 4, and 5
[7]	Mark Hopkins and Jonathan May, 2011. Tuning as Ranking. In Proceedings EMNLP 2011, ACL.	pdf
	Sections 1–4
[8a]	James Rogers, 2003. Syntactic Structures as Multi-dimensional Trees. Research on Language & Computation, 1:265–305.	pdf
	Sections 2, 4
[8b]	James Rogers, 2003. wMSO theories as grammar formalisms. Theoretical Computer Science, 293(2):291–320.	pdf
	Sections 2–4
[9]	Sebastian Seifert and Ingrid Fischer, 2004. Parsing String Generating Hypergraph Grammars. In Hartmut Ehrig, Gregor Engels, Francesco Parisi-Presicce, and Grzegorz Rozenberg, editors, Graph Transformations, volume 3256 of Lecture Notes in Computer Science, pages 263–266. Springer.	pdf
	Sections 1–6
[10a]	Laura Kallmeyer and Wolfgang Maier, 2010. Data-driven parsing with probabilistic linear context-free rewriting systems. In Proceedings of the 23rd International Conference on Computational Linguistics, COLING ’10, pages 537–545. ACL.	pdf
	Sections 2–3
[10b]	Kilian Evang and Laura Kallmeyer, 2011. PLCFRS Parsing of English Discontinuous Constituents. In Proceedings IWPT 2011, pages 104–116. ACL.	pdf
	Section 2
[11]	Shay B. Cohen and Noah A. Smith, 2010. Viterbi training for PCFGs: Hardness results and competitiveness of uniform initialization. In Proceedings ACL 2010, pages 1502–1511. ACL.	pdf
	Section 1–6

Getting Help

We have some information on writing articles available online. In general, if you have questions, do not hesitate to contact your supervisor. The earlier you address your problems, the easier the solutions will be.

TUD

Chair of Foundations of Programming