Here is a set of documents:

D1= "Lucie has a pencil"

D2= "The house is red"

D3= "The red pencil is in the red house"

D4= "The red policeman is on leave with his red shoes"

D5= "After he left the house, the policeman left on his red bicycle"

For each word appearing in the documents, calculate the idf factor.

1- The same document D was introduced x times, by mistake, in a set of documents indexed by the vector model tf.idf. Assuming x>1, comment on the effect of this error? Is the document in question more or less likely to be found among the results of a keyword search when x is large?

2- In the tf.idf vector model, what is the best strategy to find a specific document by searching by keywords:

- Choose, as keywords, the most frequent words of the document sought;

- Choose, as keywords, the most frequent words of the document sought, but which are also frequent in all the documents;

- Choose, as keywords, the most frequent words of the document sought, but which are also infrequent in all the documents.