com.bericotech.clavin.resolver.lucene
Class LuceneLocationResolver

java.lang.Object
  com.bericotech.clavin.resolver.lucene.LuceneLocationResolver

All Implemented Interfaces:

LocationResolver

public class LuceneLocationResolver
extends Object
implements LocationResolver

Resolves location names into GeoName objects. Takes location names extracted from unstructured text documents by LocationExtractor and resolves them into the appropriate geographic entities (as intended by the document's author based on context) by finding the best match in a gazetteer.

Field Summary

static org.slf4j.Logger

logger

Constructor Summary

LuceneLocationResolver(File indexDir, int maxHitDepth, int maxContextWindow)
Builds a LuceneLocationResolver by loading a pre-built Lucene index from disk and setting configuration parameters for resolving location names to GeoName objects.

Method Summary

List<ResolvedLocation>

resolveLocations(List<LocationOccurrence> locations, boolean fuzzy) Resolves the supplied list of location names into ResolvedLocations containing GeoName objects.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

logger

public static final org.slf4j.Logger logger

Constructor Detail

LuceneLocationResolver

public LuceneLocationResolver(File indexDir,
                              int maxHitDepth,
                              int maxContextWindow)
                       throws IOException,
                              org.apache.lucene.queryparser.classic.ParseException

Builds a LuceneLocationResolver by loading a pre-built Lucene index from disk and setting configuration parameters for resolving location names to GeoName objects.

Parameters:

indexDir - Lucene index directory to be loaded

maxHitDepth - number of candidate matches to consider

maxContextWindow - how much context to consider when resolving

Throws:

IOException

org.apache.lucene.queryparser.classic.ParseException

Method Detail

resolveLocations

public List<ResolvedLocation> resolveLocations(List<LocationOccurrence> locations,
                                               boolean fuzzy)
                                        throws IOException,
                                               org.apache.lucene.queryparser.classic.ParseException

Resolves the supplied list of location names into ResolvedLocations containing GeoName objects. Calls getCandidateMatches(com.bericotech.clavin.extractor.LocationOccurrence, boolean) on each location name to find all possible matches, then uses heuristics to select the best match for each by calling pickBestCandidates(java.util.List>).

Specified by:

resolveLocations in interface LocationResolver

Parameters:

locations - list of location names to be resolved

fuzzy - switch for turning on/off fuzzy matching

Returns:

list of ResolvedLocation objects

Throws:

org.apache.lucene.queryparser.classic.ParseException

IOException