public void init() {
HashMap<String, Object> config = new HashMap<String, Object>();
config.put("DF_MIN", Constants.DF_MIN);
config.put("DF_MAX", Constants.DF_MAX);
config.put("THERSAUS", ""); // 同义词库
config.put("WEIGHT", "Lucene");
config.put("CORPUS", root1.getPath());
config.put("ENCODE_TRAIN", "utf-8");
config.put("ENCODE_RECOGNIZE", "utf-8");
VSMService.build(config, new FileTokenizer());
service = VSMService.getInstance();
// _dimension();
// config.put("CORPUS", root.getPath());
// VSMService.build(config, new FileTokenizer());
// service = VSMService.getInstance();
}
private Doc buildDoc(int x, File file) {
double minScore = 0.05;
String s = FileUtil.read(file);
String[] arr = s.split("\n");
WVTWordVector vector = service.getNativeVector(s);
if (filter) {
Map<String, Double> newMap = new HashMap<String, Double>();
for (Map.Entry<String, Double> entry : vector.getTFIDFValues().entrySet()) {
if (entry.getValue() < minScore) {
newMap.put(entry.getKey(), 0D);
}
}
double[] values = vector.getValues();
for (int i = 0; i < values.length; i++) {
if (values[i] < minScore) {
values[i] = 0D;
}
}
}
return new Doc(vector, file, x, arr[2], arr[0], arr[1]);
}