我正在尝试构建一个算法来查看哪些词在评论中最常出现。
因此我想出了这个(用Javascript):
var analyze = function(comments){
var detectedWords = [];
var result = {};
comments.forEach(function(comment){
var words = comment.message.split(" ");
words.forEach(function(word){
word = word.toLowerCase();
if(word !== ""){
if(detectedWords.indexOf(word) === -1){
detectedWords.push(word);
result[detectedWords.indexOf(word)] = {"name":word,"count":1};
}else{
result[detectedWords.indexOf(word)].count++;
}
}
});
});
return _.orderBy(result, ['count'], ['desc']);
}
算法可以进一步优化吗? (内循环外的toLowerCase()?
在下一步中,我将定义一个"黑名单"或者像#34这样没有意思的词;是,我,我,是......,"
答案 0 :(得分:1)
您可以使用哈希表来引用数组项,以便更快地访问count对象。 result
现在是一个数组,现在可以排序。
var analyze = function (comments) {
var result = [],
hash = {};
comments.forEach(function (comment) {
var words = comment.message.split(" ");
words.forEach(function (word) {
word = word.toLowerCase();
if (word !== "") {
if (!hash[word]) {
hash[word] = { name: word, count: 0 };
result.push(hash[word]);
}
hash[word].count++;
}
});
});
return result.sort(function (a, b) { return b.count - a.count;});
//return _.orderBy(result, ['count'], ['desc']);
}
console.log(analyze([{ message: 'a b c d a v d e f g q' }]));