3. Assume that the total number of documents in a corpus is 128 and that the following words occur in the following number of documents:

Computer occurs in 32 documents

software occurs in 8 documents

intelligent occurs in 16 documents

robust occurs in 128 documents

1) Calculate the TF-IDF weighted term vector (WT_D = TF_D x IDF) for the following document D:

Computer intelligent software robust computer software

(Hint: all the numbers above are powers of 2, and the log in the IDF weight is taken to the base 2).

2) Suppose one has a query Q specified as:

intelligent software

Assuming that query vector is computed just in terms of TF weights (no IDF weights), and similarity is measured by the cosine metric, what is the similarity between Q and D?