/**
* Generates a List with the most common terms of the tweetsOfEvent that belong to the specific
* event. <br>
* More formally, it parses every single tweet of the event, tokenizes it and stores the terms in
* a HashMap with their respective occurencies as values.
*/
private void generateCommonTerms() {
HashMap<String, Integer> unsortedTokens = new HashMap<>();
tweetsOfEvent
.stream()
.forEach(
(tweet) -> {
String text = tweet.getText();
Tokenizer tokens =
new Tokenizer(
corpus.getConfigHandler(),
text,
corpus
.getStopWordsHandlers()
.getSWHandlerAccordingToLanguage(
LangUtils.getLangISOFromString(tweet.getLanguage())));
stemsHandler
.getStemsAsList(
tokens.getCleanTokensAndHashtags(),
Stemmers.getStemmer(LangUtils.getLangISOFromString(tweet.getLanguage())))
.stream()
.forEach(
(token) -> {
if (unsortedTokens.containsKey(token)) {
unsortedTokens.put(token, unsortedTokens.get(token) + 1); // Count it
} else {
unsortedTokens.put(token, 1);
}
});
});
allTerms = new HashSet<>(unsortedTokens.keySet());
sortMapByValue(unsortedTokens, stemsHandler);
}
/**
* Auxiliary method to sort a Map by value.
*
* @param unsortedMap The Map to be sorted.
* @param stemHandler A StemUtils object.
*/
public final void sortMapByValue(HashMap<String, Integer> unsortedMap, StemUtils stemHandler) {
// Initialize variables
Entry<String, Integer> entry;
String currentKey;
int currentValue;
// Get the 5 greatest tokens by value
// If the HashMap has less than 5 elements, just sort them
int size = (unsortedMap.keySet().size() < 5 ? unsortedMap.size() : 5);
for (int i = 0; i < size; i++) {
entry = unsortedMap.entrySet().iterator().next();
currentKey = entry.getKey();
currentValue = entry.getValue();
for (String key : unsortedMap.keySet()) {
if (unsortedMap.get(key) > currentValue) {
currentValue = unsortedMap.get(key);
currentKey = key;
}
}
commonTerms.add(stemHandler.getOriginalWord(currentKey));
unsortedMap.remove(currentKey);
}
}
