public HashSet<String> getEntitiesinTweet(String tweet) {
HashSet<String> entities = new HashSet<String>();
TwitterTokenizer tweetTokenizer = new TwitterTokenizer();
for (String token : tweetTokenizer.tokenize(tweet)) {
token = token.trim();
token =
token.replaceAll("( [^a-zA-Z0-9\\.]) | ( [^a-zA-Z0-9\\.] ) | ([^a-zA-Z0-9\\.] )", " ");
try {
Pattern p = Pattern.compile("^[A-Z]+.*");
String[] split = token.split("\\s+");
for (String s : split) {
s = s.trim();
if (p.matcher(s).matches() && !stopWords.contains(s.toLowerCase())) {
entities.add(s);
}
}
} catch (Exception e) {
e.printStackTrace();
}
for (String np : npe.extract(token)) {
if (!stopWords.contains(np.trim().toLowerCase())) {
entities.add(np.trim());
}
}
}
return entities;
}
public HashMap<String, Entity> getAllEntities(String handle) {
HashMap<String, Entity> allEntities = new HashMap<String, Entity>();
try {
BufferedReader br = new BufferedReader(new FileReader("data/" + handle + ".txt"));
BufferedWriter bw = new BufferedWriter(new FileWriter("data/" + handle + "_entities.txt"));
BufferedWriter bw1 = new BufferedWriter(new FileWriter("data/" + handle + "_statistics.txt"));
String line = "";
Counter<String> nPhraseCounter = new Counter<String>();
Counter<String> capitalsCounter = new Counter<String>();
while ((line = br.readLine()) != null) {
line = line.replaceAll("RT", "");
TwitterTokenizer tweetTokenizer = new TwitterTokenizer();
for (String token : tweetTokenizer.tokenize(line)) {
token = token.trim();
token =
token.replaceAll(
"( [^a-zA-Z0-9\\.]) | ( [^a-zA-Z0-9\\.] ) | ([^a-zA-Z0-9\\.] )", " ");
ArrayList<String> nPhrases = new ArrayList<String>();
HashSet<String> capitalWords = new HashSet<String>();
try {
Pattern p = Pattern.compile("^[A-Z]+.*");
String[] split = token.split("\\s+");
for (String s : split) {
if (p.matcher(s).matches() && !stopWords.contains(s.toLowerCase())) {
capitalWords.add(s.toLowerCase());
capitalsCounter.incrementCount(s.toLowerCase(), 1.0);
if (allEntities.containsKey(s.trim())) {
Entity e = allEntities.get(s.trim());
if (!e.tweets.contains(line)) {
e.tweets.add(line);
allEntities.put(s.trim(), e);
}
} else {
Entity e = new Entity(s.trim());
e.tweets.add(line);
allEntities.put(s.trim(), e);
}
}
}
} catch (Exception e) {
e.printStackTrace();
}
bw.write("===============================================\n");
bw.write(token + "\n");
System.out.println("token: " + token);
for (String np : npe.extract(token)) {
if (!stopWords.contains(np.trim().toLowerCase())) {
nPhrases.add(np.trim());
nPhraseCounter.incrementCount(np.trim(), 1.0);
if (allEntities.containsKey(np.trim())) {
Entity e = allEntities.get(np.trim());
if (!e.tweets.contains(line)) {
e.tweets.add(line);
allEntities.put(np.trim(), e);
}
} else {
Entity e = new Entity(np.trim());
e.tweets.add(line);
allEntities.put(np.trim(), e);
}
}
}
bw.write("===============================================\n");
bw.write("Noun-Phrases: " + nPhrases.toString() + "\n");
// HashSet<String> capitalWords =
// getCapitalizedWords(token);
if (capitalWords == null) {
bw.write("No capitals\n\n");
} else {
bw.write("Capitals: " + capitalWords.toString() + "\n\n");
}
}
bw.flush();
if (true) continue;
}
PriorityQueue<String> nPhraseQueue = nPhraseCounter.asPriorityQueue();
PriorityQueue<String> capitalQueue = capitalsCounter.asPriorityQueue();
while (nPhraseQueue.hasNext()) {
String np = nPhraseQueue.next();
bw1.write(np + " " + nPhraseCounter.getCount(np) + "\n");
}
bw1.write("=========================================================\n");
while (capitalQueue.hasNext()) {
String cap = capitalQueue.next();
bw1.write(cap + " " + capitalsCounter.getCount(cap) + "\n");
}
bw1.flush();
} catch (Exception e) {
e.printStackTrace();
}
return allEntities;
}
