[ Skip to the content ]

Institute of Formal and Applied Linguistics Wiki

[ Back to the navigation ]

You are here: start » courses » rg » 2014 » mdsm

Differences

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

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
courses:rg:2014:mdsm [2014/11/25 21:25]
nguyenti 		courses:rg:2014:mdsm [2014/11/29 13:52] (current)
popel
Line 1: Line 1:
-1. Recall from the paper presented by Tam 3 week ago.  +You should focus on the first paper (you can skip section 2.3): [[http://www.aclweb.org/anthology/W11-2503.pdf|Distributional semantics from text and images]]. 
-   a. What is the difference between distributional semantic and distributed semantic representation? +The second paper [[http://www.aclweb.org/anthology/P12-1015.pdf|Distributional Semantics in Technicolor]], an extent of the first one, is optional reading.
-   b. What is maximum size of word vector in distributional representation approach?+
  
-2. a. What is the vector representation of word "Moon" and "Sun" from co-occurrence matrix below: 
  
-       | planet | night | full | shadow | shine       +Q1.  
 +Recall the paper about word representations presented by Tam on November 10. 
 +Read http://www.quora.com/Whats-the-difference-between-distributed-and-distributional-semantic-representations 
 + 
 +(M_{w,d} is a matrix with w rows and d columns). 
 +What does w, d and k mean? 
 +What are the values of w, d and k used in the experiments in this paper? 
 + 
 +Q2. 
 +a) Compute the similarity between two words "Moon" and "Mars" from the co-occurrence matrix below. 
 +Use these raw counts (no Local Mutual Information, no normalization) and cosine similarity. 
 + 
 +           | planet | night | full | shadow | shine       
     Moon   |   34   |   27  |  19  |   9    |   20     Moon   |   34   |   27  |  19  |   9    |   20
     Sun    |   32   |   23  |  10  |   47   |   15     Sun    |   32   |   23  |  10  |   47   |   15
     Dog    |   0    |   19  |  2   |   11   |   1     Dog    |   0    |   19  |  2   |   11   |   1
     Mars   |   44   |   23  |  17  |   3    |   9     Mars   |   44   |   23  |  17  |   3    |   9
 +    
 +b) How do they deal with high dimension of vectors in those papers?
 +Can you suggest some (other) techniques of preprocessing vectors with high dimensions?
            
-        +Q3.  
-        +a) What are Bag of Word (BOVW) and Bag of Visual Word (BOW)? 
-        +b) How do they apply BOVW to compute representation of a word (concept) from a large set of images? 
-         +    
-        +Q4 (bonus). 
-        +When they construct text-based vectors of words from DM model 
- +they mentioned Local Mutual Information score. (section 3.2, also section 2.1 in the 2nd paper) 
- +So what is that score? Why did they use it?
-3. a. What are Bag of Word (BOVW) and Bag of Visual Word (BOW)? Are they synonyms? +
-   b. How do they apply BOVW to compute representation of a word (concept) +
-   from a large set of Images?+
  
 +Q5 (bonus).
 +Have you ever wished to see beautiful "Mermaids"?
 +Have you ever seen "Unicorns" in the real life?
 +"Assume that there are no photos of them on the Internet"
  
-5. Have you ever wish to see a beautiful Mermaids. +Think about a computational way to show that how they look like?
-Have you ever seen "Unicorns" in the real lie. +
-Can you think a computational way to show that how they look like?+
  
  

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