Syntactic Analysis: Literature seminars
During the literature seminars we will discuss the given scientific articles, and questions related to them. You are supposed to prepare according to the guidelines below. Teh seminars are meant as learning opportunities, where you will get help to understand two scientific articles related to the course in detail.
Groups
The seminars will be held in smaller groups, which will be announced later.
Preparation
For both seminars you are expected to read the assigned article carefully. You should also prepare answers to the questions and discussion points below and bring them to the seminar.
- Seminar 1: Mark Johnson. PCFG Models of Linguistic Tree Representations. Computational Linguistics 24(4). Pages 613-632. Also read J&M 14.9 briefly. Bachelor students do not need to go into any detail about the math, just strive to understand the reasoning behind it.
- The seminar will be held in small groups in 4-2007, Tuesday November 29, 2016.
- Group A: 13.15-14 (Evelyn, James, Jenny, Mitja Tim)
- Group B: 14.15-15 (Kajsa, Mikael, Rebecka, Rickard, Stina)
- Questions and discussion points:
- What do you think is the most important point the author makes in the article?
- What do YOU personally think is the most interesting point in the article?
- Discuss the relation between the theoretical results in section 4 and the empirical results in section 5.
- What did you learn when reading the article? (If you learnt a lot, pick a few important points, if you learnt nothing, discuss why)
- Is there something you do not understand?
- Are there any further issues you'd like to discuss?
- Seminar 2: Joakim Nivre and Jens Nilsson. Pseudo-Projective Dependency Parsing. Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL'05). Pages 99-106. Ann Arbor, USA.
I also recommend that you read the section on projectivity in the course book, 2.1.2. Section 3.5 in the course book also describes pseudo-projective parsing.
- The seminar will be held in small groups in 9-2029, Thursday January 12, 2017:
- Group D: 10.15-11 (Kajsa, Stina, James, Jenny)
- Group C: 13.15-14 (Evelyn, Tim, Rebecka, Rickard)
- Questions:
- What do you think is the most interesting point in the article?
- Discuss the relation between the results on treebank transformation, experiment 1, and parsing, experiment 2, and also the difference between the parsing results in tables 5 and 6.
- Discuss how the results on the different metrics vary on the parsing task. Do you think the metrics are well chosen? Would you have liked to see some other metric or evaluation?
- How important do you think it is to parse non-projective structures in Swedish (or English or your own language)? Consider for instance the following two Swedish sentences (somewhat simplified from the devset in assignment 4, you can look there if you want to see a non-projective analysis):
Vad kommer resultatet att bli?
Men någon utbildning var det inte möjlighet till.
Or these English sentences:
This is the book I told you about.
John saw a man yesterday with a bag.
Do you think it is reasonable to give them a non-projective analysis (as they have in the treebank for Swedish)? What causes the non-projectivity in these sentences? Can you think of other possible non-projective structures in Swedish/English/your native language.
- Do you think it seems like a good idea to treat non-projectivity by using pre- and postprocessing, or would it be better to use a parsing algorithm that can give non-projective structures directly? Discuss some pros and cons.
- What did you learn when reading the article? (If you learnt a lot, pick a few important points, if you learnt nothing, discuss why)
- Is there something you do not understand?
- Are there any further issues you'd like to discuss?
Grading
The two seminars are obligatory. In order to get a passing grade on the seminars you need to read the articles, prepare according to the instructions, and be active during the seminars. It should be clear that you have read and analysed the article, and prepared the questions, but it is perfectly fine if you found the articles difficult, did not understand everything, or if you have misunderstood some parts. If you miss a seminar, or if you are unprepared, you will have to do a complementary task. Contact Sara if this is the case.
- The seminar will be held in small groups in 4-2007, Tuesday November 29, 2016.
- Group A: 13.15-14 (Evelyn, James, Jenny, Mitja Tim)
- Group B: 14.15-15 (Kajsa, Mikael, Rebecka, Rickard, Stina)
- Questions and discussion points:
- What do you think is the most important point the author makes in the article?
- What do YOU personally think is the most interesting point in the article?
- Discuss the relation between the theoretical results in section 4 and the empirical results in section 5.
- What did you learn when reading the article? (If you learnt a lot, pick a few important points, if you learnt nothing, discuss why)
- Is there something you do not understand?
- Are there any further issues you'd like to discuss?
I also recommend that you read the section on projectivity in the course book, 2.1.2. Section 3.5 in the course book also describes pseudo-projective parsing.
- The seminar will be held in small groups in 9-2029, Thursday January 12, 2017:
- Group D: 10.15-11 (Kajsa, Stina, James, Jenny)
- Group C: 13.15-14 (Evelyn, Tim, Rebecka, Rickard)
- Questions:
- What do you think is the most interesting point in the article?
- Discuss the relation between the results on treebank transformation, experiment 1, and parsing, experiment 2, and also the difference between the parsing results in tables 5 and 6.
- Discuss how the results on the different metrics vary on the parsing task. Do you think the metrics are well chosen? Would you have liked to see some other metric or evaluation?
- How important do you think it is to parse non-projective structures in Swedish (or English or your own language)? Consider for instance the following two Swedish sentences (somewhat simplified from the devset in assignment 4, you can look there if you want to see a non-projective analysis):
Vad kommer resultatet att bli? Men någon utbildning var det inte möjlighet till.
Or these English sentences:This is the book I told you about. John saw a man yesterday with a bag.
Do you think it is reasonable to give them a non-projective analysis (as they have in the treebank for Swedish)? What causes the non-projectivity in these sentences? Can you think of other possible non-projective structures in Swedish/English/your native language. - Do you think it seems like a good idea to treat non-projectivity by using pre- and postprocessing, or would it be better to use a parsing algorithm that can give non-projective structures directly? Discuss some pros and cons.
- What did you learn when reading the article? (If you learnt a lot, pick a few important points, if you learnt nothing, discuss why)
- Is there something you do not understand?
- Are there any further issues you'd like to discuss?