Question

How to find keyword matching between multiple text blocks in PHP

Answer 1

/*
    Tokenize text into words.
    - Keeps only letters and digits
    - Splits on punctuation and spaces
*/
function tokenize(string $text): array {
    $words = [];
    $word = "";

    foreach (str_split($text) as $c) {
        if (ctype_alnum($c)) {
            $word .= strtolower($c);
        } else if ($word !== "") {
            $words[$word] = true;   // simulate std::set
            $word = "";
        }
    }

    if ($word !== "") {
        $words[$word] = true;
    }

    return $words;
}

/*
    // Find keyword matches across THREE OR MORE texts
    // -------------------------------------------------------------
    This function receives an array of sets.
    It returns the intersection of ALL sets.
*/
function findMatchesMultiple(array $allSets): array {
    if (empty($allSets)) return [];

    // Start with the first set
    $result = $allSets[0];

    // Intersect with each remaining set
    for ($i = 1; $i < count($allSets); $i++) {
        $temp = [];

        foreach ($result as $w => $_) {
            if (isset($allSets[$i][$w])) {
                $temp[$w] = true;
            }
        }

        $result = $temp;
    }

    return $result;
}

// -------------------------------------------------------------
// Three text blocks to compare
// -------------------------------------------------------------
$text1 =
    "Machine learning allows computers to learn from data. " .
    "It is widely used in modern applications.";

$text2 =
    "Data science uses machine learning techniques. " .
    "Applications rely on data-driven models.";

$text3 =
    "Modern applications of machine learning include data analysis, " .
    "automation, and intelligent systems.";

// -------------------------------------------------------------
// Tokenize all texts
// -------------------------------------------------------------
$words1 = tokenize($text1);
$words2 = tokenize($text2);
$words3 = tokenize($text3);

// Put them into an array for multi-text comparison
$allSets = [ $words1, $words2, $words3 ];

// -------------------------------------------------------------
// Find keyword matches across ALL texts
// -------------------------------------------------------------
$matches = findMatchesMultiple($allSets);

// -------------------------------------------------------------
// Output results
// -------------------------------------------------------------
echo "Matched Keywords Across ALL Texts:\n";
foreach ($matches as $w => $_) {
    echo $w . " ";
}



/*
run:

Matched Keywords Across ALL Texts:
machine learning data applications 

*/

 

70+ SQL courses for beginners and professionals