java - 使用 HashMap 计算文档频率 [Java]

Question

我正在尝试计算文档频率(即每个单词出现在多少个文档中)，例如:

Doc1:这款手机是有史以来最伟大的手机。
Doc2:您的电话号码是多少。

结果:

this              1
phone             2
is                1
the               1
ever              1
what's            1
your              1
number            1

我有以下 Java 代码

HashMap<String, String> wordDoc = new HashMap<String, String>();
HashMap<String, Integer> countDfIndex = new HashMap<String, Integer>();

if (!wordDoc.containsKey(word)) {
    wordDoc.put(word,docno);
    countDfIndex.put(word, 1);
}
if (wordDoc.get(word)!=null) {
    if(!wordDoc.containsValue(docno)) {
        wordDoc.put(word,docno);
        countDfIndex.put(word, countDfIndex.get(word)+1);
    }
}

我没有得到正确的结果，请帮忙!!

Answer 1

我假设您正在尝试计算包含相应单词的文档数量，而不是出现的总数。

如果是这样:

Map<String, Integer> countDfIndex = new HashMap<String, Integer>();

for (... document : documents) {
    Set<String> alreadyAdded = new HashSet<String>(); // new empty set for each document

    ...

    if (!alreadyAdded.contains(word)) {
        if (!countDfIndex.containsKey(word) {
            countDfIndex.put(word, 1);
        } else {
            countDfIndex.put(word, countDfIndex.get(word) + 1);
        }
        alreadyAdded.add(word); // don't add the word anymore if found again in the document
    }

}