private List<String> getFeatureValues(Instance<String> instance) {
List<String> values = new ArrayList<String>();
for (Feature feature : instance.getFeatures()) {
Object value = feature == null ? null : feature.getValue();
values.add(value == null ? null : value.toString());
}
return values;
}
public boolean encodes(Feature feature) {
if (!(feature.getValue() instanceof Counts)) return false;
Counts counts = (Counts) feature.getValue();
if (identifier == null) return true;
if (identifier.equals(counts.getIdentifier())) return true;
return false;
}
public List<NameNumber> encode(Feature feature) {
List<NameNumber> fves = new ArrayList<NameNumber>();
Counts frequencies = (Counts) feature.getValue();
String prefix = frequencies.getFeatureName();
for (Object key : frequencies.getValues()) {
String name = Feature.createName(prefix, key.toString());
NameNumber fve = new NameNumber(name, frequencies.getCount(key));
fves.add(fve);
}
normalizer.normalize(fves);
return fves;
}
@Override
public List<Feature> apply(Feature feature) {
if (i == 0) {
BufferedReader br;
try {
NERReader reader = new NERReader();
br = (BufferedReader) reader.getReader("topicCluster50.txt");
String input;
while ((input = br.readLine()) != null) {
String[] sep = input.split("\\t");
topicClass1.put(sep[0], sep[1]);
}
br.close();
} catch (UnsupportedEncodingException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (FileNotFoundException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
// System.out.println("Size:"+simWord1.size());
i++;
}
Object featureValue = feature.getValue();
if (featureValue == null) {
return Collections.singletonList(new Feature("TopicClass50", "NA"));
}
String value = featureValue.toString();
if (value == null || value.length() == 0) {
return Collections.singletonList(new Feature("TopicClass50", "NA"));
}
String output;
output = topicClass1.get(value);
// System.out.println("Size:"+i);
if (output != null) {
return Collections.singletonList(new Feature("TopicClass50", output));
}
return Collections.singletonList(new Feature("TopicClass50", "NA"));
}