You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@opennlp.apache.org by bg...@apache.org on 2013/04/04 19:10:25 UTC
svn commit: r1464643 - in
/opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps:
GeneratedSentenceProcessor.java RelatedSentenceFinder.java
SearchResultsProcessor.java
Author: bgalitsky
Date: Thu Apr 4 17:10:24 2013
New Revision: 1464643
URL: http://svn.apache.org/r1464643
Log:
opennlp-548 bing api
Modified:
opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/GeneratedSentenceProcessor.java
opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/RelatedSentenceFinder.java
opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/SearchResultsProcessor.java
Modified: opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/GeneratedSentenceProcessor.java
URL: http://svn.apache.org/viewvc/opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/GeneratedSentenceProcessor.java?rev=1464643&r1=1464642&r2=1464643&view=diff
==============================================================================
--- opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/GeneratedSentenceProcessor.java (original)
+++ opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/GeneratedSentenceProcessor.java Thu Apr 4 17:10:24 2013
@@ -84,10 +84,12 @@ public class GeneratedSentenceProcessor
||sentTry.indexOf( "mon-fri")>-1 ||sentTry.indexOf( "email us")>-1 ||sentTry.indexOf( "privacy pol")>-1 ||sentTry.indexOf( "back to top")>-1
||sentTry.indexOf( "click here")>-1 ||sentTry.indexOf( "for details")>-1 ||sentTry.indexOf( "assistance?")>-1 ||sentTry.indexOf( "chat live")>-1
||sentTry.indexOf( "free shipping")>-1 ||sentTry.indexOf( "company info")>-1 ||sentTry.indexOf( "satisfaction g")>-1 ||sentTry.indexOf( "contact us")>-1
-
||sentTry.startsWith( "fax") ||sentTry.startsWith( "write") || sentTry.startsWith( "email")||sentTry.indexOf( "conditions")>-1 ||sentTry.indexOf( "chat live")>-1
- ||sentTry.indexOf( "free shipping")>-1 ||sentTry.indexOf( "company info")>-1 ||sentTry.indexOf( "satisfaction g")>-1 ||sentTry.indexOf( "contact us")>-1
-
+ ||sentTry.startsWith( "we ") ||sentTry.indexOf( "the recipient")>-1 ||sentTry.indexOf( "day return")>-1 ||sentTry.indexOf( "days return")>-1
+
+ ||sentTry.startsWith( "fax") ||sentTry.indexOf( "refund it")>-1 || sentTry.indexOf( "your money")>-1
+ ||sentTry.indexOf( "cond???")>-1 ||sentTry.indexOf( "purchase orders")>-1
+ ||sentTry.startsWith( "exchange it ") ||sentTry.indexOf( "return it")>-1 ||sentTry.indexOf( "day return")>-1 ||sentTry.indexOf( "days return")>-1
)
return null;
Modified: opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/RelatedSentenceFinder.java
URL: http://svn.apache.org/viewvc/opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/RelatedSentenceFinder.java?rev=1464643&r1=1464642&r2=1464643&view=diff
==============================================================================
--- opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/RelatedSentenceFinder.java (original)
+++ opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/RelatedSentenceFinder.java Thu Apr 4 17:10:24 2013
@@ -69,7 +69,7 @@ public class RelatedSentenceFinder {
public List<HitBase> findRelatedOpinionsForSentenceFastAndDummy(String word,
List<String> sents) throws Exception {
BingWebQueryRunner yrunner = new BingWebQueryRunner();
- List<HitBase> searchResult = yrunner.runSearch(word);
+ List<HitBase> searchResult = yrunner.runSearch(word, 100);
return searchResult;
}
@@ -83,7 +83,7 @@ public class RelatedSentenceFinder {
for (String query : nounPhraseQueries) {
System.out.println("\nquery = " + query);
// query += " "+join(MENTAL_VERBS, " OR ") ;
- List<HitBase> searchResult = yrunner.runSearch(query);
+ List<HitBase> searchResult = yrunner.runSearch(query, 100);
if (searchResult != null) {
for (HitBase item : searchResult) { // got some text from .html
if (item.getAbstractText() != null
@@ -123,7 +123,7 @@ public class RelatedSentenceFinder {
BingWebQueryRunner yrunner = new BingWebQueryRunner();
for (String verbAddition : StoryDiscourseNavigator.frequentPerformingVerbs) {
List<HitBase> searchResult = yrunner.runSearch(sentence + " "
- + verbAddition);
+ + verbAddition, 100);
if (searchResult != null) {
for (HitBase item : searchResult) { // got some text from .html
if (item.getAbstractText() != null
Modified: opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/SearchResultsProcessor.java
URL: http://svn.apache.org/viewvc/opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/SearchResultsProcessor.java?rev=1464643&r1=1464642&r2=1464643&view=diff
==============================================================================
--- opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/SearchResultsProcessor.java (original)
+++ opennlp/sandbox/opennlp-similarity/src/main/java/opennlp/tools/similarity/apps/SearchResultsProcessor.java Thu Apr 4 17:10:24 2013
@@ -90,24 +90,17 @@ public class SearchResultsProcessor exte
}
public List<HitBase> runSearchViaAPI(String query) {
- BingResponse resp = null, // obtained from bing
- newResp = null; // re-sorted based on similarity
+ List<HitBase> hits = null;
try {
- List<String> resultList = search(query, "", "", 30);
- resp = populateBingHit(resultList.get(0));
+ List<HitBase> resultList = runSearch(query, 30);
// now we apply our own relevance filter
- newResp.setHits(calculateMatchScoreResortHits(resp.getHits(), query));
+ hits = calculateMatchScoreResortHits(resultList, query);
} catch (Exception e) {
// e.printStackTrace();
LOG.info("No search results for query '" + query);
- e.printStackTrace();
return null;
}
- // cast to super class
- List<HitBase> hits = new ArrayList<HitBase>();
- for (HitBase h : resp.getHits())
- hits.add((HitBase) h);
hits = removeDuplicates(hits, 0.9);