You are viewing a plain text version of this content. The canonical link for it is here.
Posted to java-user@lucene.apache.org by Ranjit Kumar <Ra...@otssolutions.com> on 2010/12/25 06:46:07 UTC
Lucene: how to get frequency of Boolean query
Hi,
Merry Christmas!!
In case of Boolean query like 'sql AND server' .
I am using parser to get correct document containing both sql and server. Inside for loop in below code I get correct documented and to get frequency I need to sum frequency of 'sql' and 'server' individually with the help of termDocs.read().
As I am searching through millions of document. So, to calculate frequency it takes about 160 Second for 80k document.
Is there any way to get frequency of 'Boolean query' directly without manipulation. As it takes lots of time. In case of single term and phrase query, I got frequency for 80k document within 10 Seconds.
QueryParser parser = new QueryParser(Version.LUCENE_CURRENT, field, analyzer);
parser.setDefaultOperator(QueryParser.AND_OPERATOR);
Query query = parser.parse("sql AND server");
TopDocs docs = searcher.search(query, null, n);
TermDocs termDocs = reader.termDocs();
termDocs.seek(new Term(field, query.toString(field).split(" ")[0].toLowerCase()));
int[] ints = new int[docs.totalHits];
int[] ints1 = new int[docs.totalHits];
termDocs.read(ints, ints1);
List lsts = Arrays.asList(ArrayUtils.toObject(ints));
List lsts1 = Arrays.asList(ArrayUtils.toObject(ints1));
termDocs.seek(new Term(field, query.toString(field).split(" ")[1].toLowerCase()));
int[] inta = new int[docs.totalHits];
int[] inta1 = new int[docs.totalHits];
termDocs.read(inta, inta1);
List lsta = Arrays.asList(ArrayUtils.toObject(inta));
List lsta1 = Arrays.asList(ArrayUtils.toObject(inta1));
int totalFreq = 0;
int docId = -1;
String a = null;
int contId = 0;
String path=null;
for (int i = 0; i < docs2.scoreDocs.length; i++) {
docId = docs2.scoreDocs[i].doc;
path=reader.document(docId).get("path");
a = path.substring(path.lastIndexOf("\\") + 1, path.lastIndexOf("."));
try {
if(a.indexOf("e")>-1){
contId = Integer.parseInt(a.substring(0, a.length()-1));
}else{
contId = Integer.parseInt(a);
}
} catch (Exception e) {
// e.printStackTrace();
}
if ((lsts.indexOf(docId) > -1 && lsta.indexOf(docId) > -1) || (lsta.indexOf(docId) > -1) && lsts.indexOf(docId) > -1) {
totalFreq = (Integer) lsts1.get(lsts.indexOf(docId)) + (Integer) lsta1.get(lsta.indexOf(docId));
} else if (lsts.indexOf(docId) > -1) {
totalFreq = (Integer) lsts1.get(lsts.indexOf(docId));
} else if (lsta.indexOf(docId) > -1) {
totalFreq = (Integer) lsta1.get(lsta.indexOf(docId));
}
w.write(contId+"\t"+ID+"\t"+totalFreq+"\t"+reader.document(docId).get("path")+"\n");
}
Thanks & Regards,
Ranjit Kumar
Associate Software Engineer
[cid:image002.jpg@01CB7089.C0069B40]
US: +1 408.540.0001
UK: +44 208.099.1660
India: +91 124.474.8100 | +91 124.410.1350
FAX: +1 408.516.9050
http://www.otssolutions.com
=================================================================================================== Private, Confidential and Privileged. This e-mail and any files and attachments transmitted with it are confidential and/or privileged. They are intended solely for the use of the intended recipient. The content of this e-mail and any file or attachment transmitted with it may have been changed or altered without the consent of the author. If you are not the intended recipient, please note that any review, dissemination, disclosure, alteration, printing, circulation or Transmission of this e-mail and/or any file or attachment transmitted with it, is prohibited and may be unlawful. If you have received this e-mail or any file or attachment transmitted with it in error please notify OTS Solutions at info@otssolutions.com ===================================================================================================
Re: Lucene: how to get frequency of Boolean query
Posted by Li Li <fa...@gmail.com>.
do you mean get tf of the hited documents when doing search?
it's a difficult problem because only TermScorer has TermDocs and using tf
in score() function.
but this we can't know whether these doc is selected because we use a
priorityQueue in TopScoreDocCollector
public void collect(int doc) throws IOException {
float score = scorer.score();
// This collector cannot handle these scores:
assert score != Float.NEGATIVE_INFINITY;
assert !Float.isNaN(score);
totalHits++;
if (score <= pqTop.score) {
// Since docs are returned in-order (i.e., increasing doc Id), a
document
// with equal score to pqTop.score cannot compete since HitQueue
favors
// documents with lower doc Ids. Therefore reject those docs too.
return;
}
pqTop.doc = doc + docBase;
pqTop.score = score;
pqTop = (ScoreDoc) pq.updateTop();
}
Because the scorer in BooleanScorer2 is
ConjuctionScorer(DisjuctionScorer). it can not get tf.
If you want to do this. you have to modify many codes
2010/12/25 Ranjit Kumar <Ra...@otssolutions.com>
> Hi,
>
> Merry Christmas!!
>
>
>
> In case of Boolean query like *’sql AND server’ *.
>
> I am using parser to get correct document containing both *sql* and *
> server*. Inside for loop in below code I get correct documented and to get
> frequency I need to sum frequency of ‘sql’ and ‘server’ individually with
> the help of *termDocs.read(). *
>
> As I am searching through millions of document. So, to calculate frequency
> it takes about *160 Second* for *80k* document.
>
>
>
> Is there any way to get frequency of ‘*Boolean query’ *directly without
> manipulation. As it takes lots of time. In case of single term and phrase
> query, I got frequency for 80k document within 10 Seconds.
>
>
>
> *QueryParser parser = new QueryParser(Version.LUCENE_CURRENT, field,
> analyzer);*
>
> *parser.setDefaultOperator(QueryParser.AND_OPERATOR);*
>
> *Query query = parser.parse(“sql AND server”);*
>
> *TopDocs docs = searcher.search(query, null, n);*
>
> *TermDocs termDocs = reader.termDocs();*
>
> * termDocs.seek(new Term(field,
> query.toString(field).split(" ")[0].toLowerCase()));*
>
> * int[] ints = new int[docs.totalHits];*
>
> * int[] ints1 = new int[docs.totalHits];*
>
> * termDocs.read(ints, ints1);*
>
> * List lsts = Arrays.asList(ArrayUtils.toObject(ints));
> *
>
> * List lsts1 =
> Arrays.asList(ArrayUtils.toObject(ints1));*
>
> * *
>
> * termDocs.seek(new Term(field,
> query.toString(field).split(" ")[1].toLowerCase()));*
>
> * int[] inta = new int[docs.totalHits];*
>
> * int[] inta1 = new int[docs.totalHits];*
>
> * termDocs.read(inta, inta1);*
>
> * List lsta = Arrays.asList(ArrayUtils.toObject(inta));
> *
>
> * List lsta1 =
> Arrays.asList(ArrayUtils.toObject(inta1));*
>
> * *
>
> * int totalFreq = 0;*
>
> * int docId = -1;*
>
> * String a = null;*
>
> * int contId = 0;*
>
> * String path=null;*
>
> *for (int i = 0; i < docs2.scoreDocs.length; i++) { *
>
> * docId = docs2.scoreDocs[i].doc;*
>
> * path=reader.document(docId).get("path");*
>
> * a = path.substring(path.lastIndexOf("\\") + 1,
> path.lastIndexOf("."));*
>
> * try {*
>
> * if(a.indexOf("e")>-1){*
>
> * contId =
> Integer.parseInt(a.substring(0, a.length()-1));*
>
> * }else{*
>
> * contId = Integer.parseInt(a);*
>
> * }*
>
> * } catch (Exception e) {*
>
> *// e.printStackTrace();*
>
> * }*
>
> * *
>
> * if ((lsts.indexOf(docId) > -1 &&
> lsta.indexOf(docId) > -1) || (lsta.indexOf(docId) > -1) &&
> lsts.indexOf(docId) > -1) {*
>
> * totalFreq = (Integer)
> lsts1.get(lsts.indexOf(docId)) + (Integer) lsta1.get(lsta.indexOf(docId));
> *
>
> * } else if (lsts.indexOf(docId) > -1) {*
>
> * totalFreq = (Integer)
> lsts1.get(lsts.indexOf(docId));*
>
> * } else if (lsta.indexOf(docId) > -1) {*
>
> * totalFreq = (Integer)
> lsta1.get(lsta.indexOf(docId));*
>
> * }*
>
> * *
>
> *
> w.write(contId+"\t"+ID+"\t"+totalFreq+"\t"+reader.document(docId).get("path")+"\n");
> *
>
> * }*
>
>
>
>
>
> Thanks & Regards,
>
> *Ranjit Kumar ***
>
> Associate Software Engineer
>
>
>
> [image: cid:image002.jpg@01CB7089.C0069B40]
>
>
>
> *US:* +1 408.540.0001
>
> *UK:* +44 208.099.1660
>
> *India:* +91 124.474.8100 | +91 124.410.1350
>
> *FAX:* +1 408.516.9050
>
> http://www.otssolutions.com
>
>
> ===================================================================================================
> Private, Confidential and Privileged. This e-mail and any files and
> attachments transmitted with it are confidential and/or privileged. They are
> intended solely for the use of the intended recipient. The content of this
> e-mail and any file or attachment transmitted with it may have been changed
> or altered without the consent of the author. If you are not the intended
> recipient, please note that any review, dissemination, disclosure,
> alteration, printing, circulation or Transmission of this e-mail and/or any
> file or attachment transmitted with it, is prohibited and may be unlawful.
> If you have received this e-mail or any file or attachment transmitted with
> it in error please notify OTS Solutions at info@otssolutions.com===================================================================================================
>
>