The ranker we use is QL
4.3 The script used for 4.3 has to change, otherwise it couldn't be properly executed

For the pagerank Algorithm, we tested the following four scenario:
lamda = 0.1 & iteration = 1: 0.4602262832665721
lamda = 0.1 & iteration = 2: 0.44370887860546315
lamda = 0.9 & iteration = 1: 0.46022629083686817
lamda = 0.9 & iteration = 2: 0.44358545778068725

We will choose lamda = 0.9 and iteration = 2. Because lamda = 0.1 will introduce too much noise and is not taking too much advantage of the graph. For the number of iteration, we should make it larger because it will not be stable if we only do one iteration.


The result file for pagerank and numviews is under temp/:
Pagerank: temp/pageRank.txt
numviews: temp/numviews.txt


rm -f prf*.tsv
i=0
while read q ; do
i=$((i + 1));
prfout=prf-$i.tsv;
Q=`echo $q| sed "s/ /%20/g"`;
t="http://localhost:25815/prf?query=$Q&ranker=QL&numdocs=10&numterms=100"
curl $t > $prfout;
echo $q:$prfout >> prf.tsv
done < queries.tsv
java -cp src edu.nyu.cs.cs2580.Bhattacharyya prf.tsv qsim.tsv