为高级 taxt 创建数组联合,比较像 Jaccard 索引

create array union for advanced taxt compare like Jaccard index

我正在尝试开发在 2 个数组上创建联合以比较大文本的功能。例如:

$myText1 = "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Etiam quis eleifend sem. Maecenas at elit varius erat malesuada mollis. Nullam vulputate, velit vel posuere finibus, ex quam imperdiet lorem, eu gravida quam purus et augue. Nunc dictum nunc vehicula mattis mollis. Quisque pharetra lorem id ultrices feugiat. Vestibulum ante ipsum primis in faucibus orci luctus et ultrices posuere cubilia Curae; Suspendisse sed tortor odio. Proin consequat sed purus quis congue. Donec faucibus nec magna ac sodales."
$myText2="Mauris eget metus non mi tristique varius. Ut suscipit ante id condimentum interdum. Pellentesque egestas, leo quis tincidunt ornare, nulla metus sagittis lorem, in interdum justo tellus ac lectus. Vivamus auctor bibendum eros vel cursus. Praesent semper mauris dolor, sit amet placerat orci vestibulum non. Aenean consequat ultrices massa, in congue urna condimentum nec. Pellentesque eu faucibus dolor. Ut eu accumsan nunc, vel egestas dolor. Sed convallis mi et orci interdum tincidunt."
$str1array = explode(' ', $myText1 );
$str2array = explode(' ', $myText2);
$dallas = array();
foreach($str1array as $str){
    $dallas[] = trim($str, " .,;\t\n\r[=10=]\x0B");
}
$phoenix = array();
foreach($str2array as $str){
    $phoenix[] = trim($str, " .,;\t\n\r[=10=]\x0B");
}
$inter = array_intersect($dallas, $phoenix);
$union = array_unique(array_merge($str1array, $str2array));
$indiceDeJaccard = count($inter)/count($union);
$coefficientDeDice = (2*count($inter))/(count($str1array)+count($str2array));

但是array_unique的问题是抑制了原文中的所有重复内容,并为计算Jaccard指数创建了虚假数据。

抱歉,我已经回答了我的问题:

$inter = array_intersect($dallas, $phoenix);
$diff = array_diff($dallas, $phoenix);
$union = array_merge($inter, $diff);