用node写计算长度不相同的两个文本之间的相似度的代码

计算两个文本之间的相似度可以使用余弦相似度公式来实现。以下是使用Node.js编写的示例代码：

function calculateSimilarity(text1, text2) {
  // 将文本转换为单词数组
  const words1 = text1.toLowerCase().split(/[^\w]+/).filter(Boolean);
  const words2 = text2.toLowerCase().split(/[^\w]+/).filter(Boolean);

  // 将所有单词放入一个集合中，以便计算出所有单词的数量
  const allWords = new Set([...words1, ...words2]);

  // 计算每个文本中每个单词的出现次数
  const wordCounts1 = countWords(words1, allWords);
  const wordCounts2 = countWords(words2, allWords);

  // 计算余弦相似度
  const dotProduct = calculateDotProduct(wordCounts1, wordCounts2);
  const magnitude1 = calculateMagnitude(wordCounts1);
  const magnitude2 = calculateMagnitude(wordCounts2);
  const similarity = dotProduct / (magnitude1 * magnitude2);

  return similarity;
}

// 计算单词在文本中的出现次数
function countWords(words, allWords) {
  const wordCounts = {};

  for (const word of allWords) {
    wordCounts[word] = 0;
  }

  for (const word of words) {
    wordCounts[word]++;
  }

  return wordCounts;
}

// 计算两个文本的点积
function calculateDotProduct(wordCounts1, wordCounts2) {
  let dotProduct = 0;

  for (const word in wordCounts1) {
    dotProduct += wordCounts1[word] * wordCounts2[word];
  }

  return dotProduct;
}

// 计算文本的向量大小
function calculateMagnitude(wordCounts) {
  let sumOfSquares = 0;

  for (const word in wordCounts) {
    sumOfSquares += wordCounts[word] ** 2;
  }

  return Math.sqrt(sumOfSquares);
}

使用示例：

const text1 = "Hello world! This is a test.";
const text2 = "This is a test. Hello world!";

const similarity = calculateSimilarity(text1, text2);
console.log(`The similarity between the texts is ${similarity}`);

输出结果：

The similarity between the texts is 1

请注意，此示例代码假设输入的文本为纯文本字符串。如果您的输入包含HTML标记或其他格式，则需要对文本进行预处理，以便正确提取单词。