Description
See [this document](../Practicum.md) for general information about the practicums.
Learning objectives:
– Ranking documents using the Vector Space Model
– Building an inverted index
Task 1. Term weighting and vector space retrieval
– Score a toy-sized document collection against a query using the vector space model (i.e., TFIDF term weighting and cosine similarity).
Task 2. Building an inverted index
– You are given a sample (1000 documents) from the [The Reuters-21578 data collection](http://www.daviddlewis.com/resources/testcollections/reuters21578/) in `data/reuters21578-000.xml`
– The code that parses the XML and extract a list of preprocessed terms (tokenized, lowercased, stopwords removed) is already given.
– You are also given an InvIndex class that manages the posting lists operations.
– Build an inverted index from the input collection with the term frequencies stored.
– Save the inverted index to a text file. E.g., `termID docID1:freq1 docID2:freq2 …`.
