Ruby：如何测试两个文本块之间的相似性？

文本1：

absolute obedience to the zerg collective sentience known as the Overmind. The Overmind directed the actions of every zerg creature in the Swarm,functioning through a hierarchy of lesser sentients.

文本2：

zerg creature in the Swarm,functioning through a hierarchy of lesser sentients. Although the Overmind was primarily driven by its desire to consume and assimilate

文字3

When the zerg first arrived in the Koprulu sector,they were unified by their absolute obedience to the zerg collective sentience known as the Overmind. The Overmind directed the actions of every zerg creature in the Swarm,functioning through a hierarchy of lesser sentients. Although the Overmind was primarily driven by its desire to consume and assimilate the advanced protoss race,it found useful but undeveloped material in humanity.

现在,Text1的结尾和text2的开头重叠,所以我们说文本块不是唯一的.类似地,使用Text3,Text1可以在里面找到(以及Text2),所以由于重叠,这也不是唯一的.

所以,我的问题：

如何编写可以查看连续字母或单词并确定唯一性的内容？理想情况下,我希望这样的方法返回一些值,表示相似度 – 可能是两个文本块大小的平均值匹配的单词数.当它返回0时,测试的两个文本应该是完全唯一的.

在使用Ruby的字符串方法时,我遇到了一些问题.

首先,我开始尝试找到两个字符串的交集.

>> a = "nt version,there are no ch"  
>> b = "he current versi"  
>> (a.chars.to_a & b.chars.to_a).join  
=> "nt versihc"

上述方法的问题在于它只是在结果的末尾添加了共同的字母(我们失去了字符的顺序),这将使得难以测试唯一性.但我不认为交叉是开始这种相似性比较的最佳方式.在被比较的两个文本中可以存在任意数量的单词组合.所以也许如果我创建了一系列连续的相似性……但是这需要我们在尝试短语长度时遍历其中一个文本.

我想我真的只是不知道从哪里开始,并且以一种有效而不是O(n ^ too_high)的方式.