[ Skip to the content ]

Institute of Formal and Applied Linguistics Wiki

[ Back to the navigation ]

You are here: start » courses » rg » 2013 » jacana-align

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision 		Previous revision
courses:rg:2013:jacana-align [2013/11/04 21:20]
fiala created 		courses:rg:2013:jacana-align [2013/11/04 22:22] (current)
popel
Line 2: Line 2:
 Slides accompanying the article are located at [[http://cs.jhu.edu/~xuchen/paper/yao-jacana-wordalign-acl2013.ppt]] Slides accompanying the article are located at [[http://cs.jhu.edu/~xuchen/paper/yao-jacana-wordalign-acl2013.ppt]]
  
-**1/** Describe parameters of feature function in section 3.1+**1/** Describe parameters of feature function in section 3.1. Is there something unclear in the formula?
  
-**2/** Section 3.1 describes problem with unknown number of states. +**2/** Section 3.1 describes a problem with unknown number of states. 
-For example, we have following source sentence: //"Linda is married to Charles."// +For example, we have the following source sentence: //"Linda is married to Charles."// 
-Target sentence: //"Charles likes his new role as the husband."// +and target sentence: //"Charles likes his new role as the husband."// 
-And tags: NOUN, VERB, ADJ, ADV, PRON, DET, ADP, NUM, CONJ, PRT+and tags: NOUN, VERB, ADJ, ADV, PRON, DET, ADP, NUM, CONJ, PRT
 How many states will be used? How many states will be used?
  
-**3/** Search WordNet (http://wordnetweb.princeton.edu/perl/webwn) for following words and write answers: +**3/** Search WordNet (http://wordnetweb.princeton.edu/perl/webwn) for words "married" and "husband". Which (non-zero) WordNet features will be triggered for these two words?
-**a /** Imagine, that source sentence countains word //"well"// and target //"part"//. +
-Find differences with incorporating POS tagger and without it. How big will be gain in number of comparisons (just roughly)? +
-**b/** Is there a synset containing both //"well"// and //"water"// words?+
  
-**4/** Have a look at Table 3. How big was gain from incorporating Wordnet with POS tagger? Would you use two external resources for this gain (refer discussion in section 4.6)?+**a/** using POS tagger 
 +**b/** without POS tagger
  
-**5/** Why do you think, authors add Hamming distance into first formula in section 3.1. Can normalization function work without it?+**4/** Why do you think, the authors add Hamming distance (cost) into the first formula in Section 3.1. Can the normalization function Z work without it?

[ Back to the navigation ] [ Back to the content ]