hw4 - CSE671 - Introduction to DB Systems II Instructor:...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
CSE671 - Introduction to DB Systems II Instructor: Hui Fang Homework 4: Implementing Pivoted in Lucene Solution due: in class on Feb.27, 2008 Code due: 11:59pm on Feb. 27, 2008 Note that this is a group assignment (2 students per group). Please send your group preference to TA by Feb. 16, 2008. After Feb. 16, 2008, TA will randomly pair you up if you do not have any preference. In this assignment, you will use an open source IR toolkit (i.e., Lucene). You need to implement the pivoted normalization function to replace the default similarity function in Lucene, and evaluate both retrieval functions. (1) Getting started Use ssh to remotely log onto cweb.cse.oho-state.edu Type “subscribe” and select “11” to subscribe Java. Download the homework directory at http://www.cse.ohio-state.edu/~hfang/cse671/hw/hw4.tar.gz Use the command “gtar xzvf hw4.tar.gz”. You will see a “hw4” directory after typing the above command line, and there are four folders, i.e., “hints”, “data”, “test” and “lucene-2.2.0”, under hw4 directory. Go to “hw4/lucene-2.2.0” directory Type “/opt/apache-ant-1.7.0/bin/ant” to compile lucene (2) Files under hw4 directory lucene-2.2.0: source codes of lucene. data: in this directory, you could find the document collection (database.sgml), query file (query) and judgement file (qrel). test: In this directory, you will find IndexDocs.java (to build the index), Retrieval.java (to run the retrieval function) , and ireval.pl (to evaluate the results). We implemented a simple index
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
application and a simple retrieval application with Lucene API. IndexDocs.java takes a file (whose format is similar to database.sgml ) as input and indexes all the documents in that file using Lucene. Each document in the index contains two fields: docid and content. Only content field is indexed. The generated index files are stored under index directory. Retrieval.java returns top 1000 documents indexed under index directory for each query given in the query file. To compile them, you can go to test directory and type javac –classpath . ./lucene-2.2.0/build/classes/java *.java
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 5

hw4 - CSE671 - Introduction to DB Systems II Instructor:...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online