CSCE 5200 Information Retrieval and Web Search
Spring 2007
Assignment 2
Issued: 02/21/2008 Due: 03/14/2008




The Cranfield collection is a standard IR text collection, consisting of 
1400 documents from the aerodynamics field. It is available from the class 
web page. Several queries, and relevance judgments associated with this 
query, are also provided from the class web page (under the Assignments
section)

To complete this assignment, you are encouraged to use the pre-processing
tools implemented during previous assignments.


1. Implement an indexing scheme based on the vectorial model, as discussed
in class. The steps pointed out in class can be used as guidelines for the
implementation. For weighting, use (1) the TF/IDF weighting scheme, and 
(2) select an additional weighting scheme from Salton&Buckley
paper on "Term weighting" that we discussed in class. Select a scheme
that in your opinion should lead to high retrieval efficiency. 
[which means that the criterion for deciding on a weighting scheme should 
be efficiency, rather than simplicity!]. Add this weighting scheme to your
indexing program.

2. For each of the ten queries provided on the class webpage, determine a 
ranked list of documents, in descending order of their similarity with the 
query. The output of your retrieval should be a list of (query_id, document_id)
 pairs. 

Determine the average precision and recall for the ten queries, when you use:
- top 10 documents in the ranking
- top 50 documents in the ranking
- top 100 documents in the ranking
- top 500 documents in the ranking

Repeat this experiment for the second weighting scheme you 
selected from the Salton&Buckley paper.
Which weighting scheme provides better results?

A list of relevant documents for each query is provided on the class 
webpage, such that you can determine precision and recall.


Submission instructions:
   - write a README file including:
         * a detailed note about the functionality of the above programs, 
         * complete instructions on how to run them
         * answers to the questions above 
   - make sure you include your name in each program and in the README file.
   - make sure all your programs run correctly on the CSP machines.
   - submit your assignment, including programs and README file by the
     due date using the 'project' program. 
     Class code is 5200s001, project HW2.