use of org.opensextant.extractors.xtax.TaxonMatcher in project Xponents by OpenSextant.
the class PlaceGeocoder method configure.
/**
* We do whatever is needed to init resources... that varies depending on
* the use case.
*
* Guidelines: this class is custodian of the app controller, Corpus feeder,
* and any Document instances passed into/out of the feeder.
*
* This geocoder requires a default /exclusions/person-name-filter.txt,
* which can be empty, but most often it will be a list of person names
* (which are non-place names)
*
* Rules Configured in approximate order:
*
* <pre>
* CountryRule -- tag all country names
* NameCodeRule -- parse any Name, CODE, or Name1, Name2 patterns for "Place, AdminPlace" evidence
* PersonNameRule -- annotate, negate any patterns or matches that appear to be known celebrity persons or organizations.
* Qualified places are not negated, e.g., "Euguene, Oregon" is a place; "Euguene" with no other evidence is a person name.
* CoordRule -- if requested, parse any coordinate patterns; Reverse geocode Country + Province.
* ProvinceAssociationRule -- associate places with Province inferred by coordinates.
* MajorPlaceRule -- identify major places by feature type, class or location population.
* LocationChooserRule -- final rule that assigns confidence and chooses best location(s)
*
* Your Rule Here -- use addRule( GeocodeRule ) to add a rule on the stack. It will be evaluated just before the final LocationChooserRule.
* your rule should improve Place scores on PlaceCandidates and name the rules that fire.
* </pre>
*
* @throws ConfigException
* on err
*/
@Override
public void configure() throws ConfigException {
// ==============
// Rule setup: Create GeocodeRules, add them to this.rules if they can be evaluated generically
// on a list of place tags.
// Otherwise such rules are configured, set during the request, and evaluated adhoc as you need.
//
/* assess country names and codes */
countryRule = new CountryRule();
countryRule.setCountryObserver(this);
/* assess NAME, CODE patterns */
nameWithAdminRule = new NameCodeRule();
nameWithAdminRule.setBoundaryObserver(this);
// Nonsense is filtered out, rather than scored and ranked low.
nonsenseFilter = new NonsenseFilter();
rules.add(nonsenseFilter);
/**
* Files for Place Name filter are editable, as you likely have
* different ideas of who are "person names" to exclude when they
* conflict with place names. If you are filtering out such things, then
* it makes sense to filter them out earliest and not incorporate them
* in geocoding.
*
*/
personNameRule = new PersonNameFilter("/filters/person-name-filter.txt", "/filters/person-title-filter.txt", "/filters/person-suffix-filter.txt");
rules.add(personNameRule);
/*
* assess coordinates related to ADM1, CC
*/
coordRule = new CoordinateAssociationRule();
coordRule.setCountryObserver(this);
coordRule.setLocationObserver(this);
if (xcoord == null && (isCoordExtractionEnabled())) {
xcoord = new XCoord();
xcoord.configure();
/*
* assess ADM1 related to found NAMES as a result of coordinates
*/
adm1Rule = new ProvinceAssociationRule();
adm1Rule.setCountryObserver(this);
rules.add(coordRule);
rules.add(adm1Rule);
}
//
try {
Map<String, Integer> popstats = GeonamesUtility.mapPopulationByLocation(GeonamesUtility.loadMajorCities("/geonames.org/cities15000.txt"));
majorPlaceRule = new MajorPlaceRule(popstats);
} catch (IOException err) {
log.error("Xponents 2.8: cities population data is used for geocoding. Will continue without it.");
majorPlaceRule = new MajorPlaceRule(null);
}
majorPlaceRule.setCountryObserver(this);
majorPlaceRule.setBoundaryObserver(this);
rules.add(majorPlaceRule);
//
if (isPersonNameMatchingEnabled()) {
try {
personMatcher = new TaxonMatcher();
personMatcher.configure();
/*
* Default catalog must be built. Extraction ./XTax folder has
* script for populating a catalog.
*/
personMatcher.addCatalogFilter("JRC");
personMatcher.addCatalogFilter("nationality");
} catch (IOException err) {
throw new ConfigException("XTax resource not available.");
}
}
// Un-filter city names that can be resolved if other ADMIN places line up.
// E.g., "Cleveland Caveliers" filters out Cleveland, but if Cleveland is mentioned alone
// then Cleveland itself will be promoted to a location. E.g., sports teams travel
// so mention of "Cleveland Caveliers visiting Seattle" would not geolocate this to Ohio
// unless the city or state was mentioned separately.
//
placeInOrgRule = new ContextualOrganizationRule();
placeInOrgRule.setBoundaryObserver(this);
rules.add(placeInOrgRule);
// Simple patterns such as city of x or abc county.
//
rules.add(new NameRule());
chooser = new LocationChooserRule();
chooser.setCountryObserver(this);
chooser.setBoundaryObserver(this);
chooser.setLocationObserver(this);
//rules.add(chooser);
countryCatalog = this.getGazetteer().getCountries();
}
use of org.opensextant.extractors.xtax.TaxonMatcher in project Xponents by OpenSextant.
the class TaxonomicTagger method main.
/**
* Do a basic test of TaxTagger. Read a file, tag it.
*
* @param args
*/
public static void main(String[] args) throws Exception {
gnu.getopt.Getopt opts = new gnu.getopt.Getopt("TaxonomicTagger", args, "i:c:");
int c = -1;
String file = null;
String catalog = null;
while ((c = opts.getopt()) != -1) {
switch(c) {
case 'i':
file = opts.getOptarg();
break;
case 'c':
catalog = opts.getOptarg();
break;
default:
System.out.println("Usage -i filename [-c cat]");
System.exit(-1);
}
}
taxtag = new TaxonMatcher();
taxtag.configure();
try {
if (new File(file).isDirectory()) {
File[] files = new File(file).listFiles();
for (File f : files) {
System.out.println("\n\nFile = " + f + "\n=========MATCHES:=========");
String doc = FileUtility.readFile(f);
testDoc(doc);
}
return;
}
String doc = FileUtility.readFile(file);
if (catalog != null) {
// Invalid filter + valid filter.
System.out.println("Testing a valid catalog - tagger will return only matches from subset catalog=" + catalog);
taxtag.addCatalogFilter(catalog);
}
testDoc(doc);
} catch (Exception err) {
err.printStackTrace();
}
taxtag.cleanup();
}
use of org.opensextant.extractors.xtax.TaxonMatcher in project Xponents by OpenSextant.
the class KeywordTaggerMapper method setup.
/**
* Setup. XTax or PlaceGecoder takes in SOLR path for xponents solr from JVM environment.
*/
@Override
public void setup(Context c) throws IOException {
super.setup(c);
try {
xtax = new TaxonMatcher();
} catch (ConfigException e) {
// TODO Auto-generated catch block
throw new IOException("setup.XTax", e);
}
log.info("DONE");
}
use of org.opensextant.extractors.xtax.TaxonMatcher in project Xponents by OpenSextant.
the class TestXTax method main.
/**
*
*
* @param args
* the arguments
* @throws Exception
* the exception
*/
public static void main(String[] args) {
TaxonMatcher tax = null;
try {
tax = new TaxonMatcher();
// Find JRC entities that have this random id pattern ID='1*1' and are in Russian form.
//
List<Taxon> results = tax.search("tag:jrc_id+1*1 AND tag:lang_id+ru");
for (Taxon tx : results) {
System.out.println("Found: " + getJRCTag(tx.tagset) + " = " + tx);
}
if (args.length > 0) {
File f = new File(args[0]);
String content = FileUtility.readFile(f, "UTF-8");
List<TextMatch> findings = tax.extract(content);
for (TextMatch tm : findings) {
String type = "" + ((TaxonMatch) tm).getTaxons();
System.out.println("Found: " + tm + "\n\t\t" + type);
}
}
} catch (Exception err) {
err.printStackTrace();
} finally {
tax.shutdown();
System.exit(0);
}
}
Aggregations