shopbuilder / app / AI / Embedding / DataEmbedding.php

DataEmbedding.php in ShopBuilder – WooCommerce Builder For Elementor 3.2.2, at app/AI/Embedding/DataEmbedding.php

157 lines 4.9 KB

Raw Download Zip

1	<?php
2	/**
3	* Class DataEmbedding
4	*
5	* Handles the generation, storage, and retrieval of AI-powered text embeddings
6	* for WooCommerce products using supported AI services like OpenAI, Gemini,
7	* or DeepSeek. Also provides semantic search capabilities using cosine similarity.
8	*
9	* @package RadiusTheme\SB\AI\Embedding
10	* @since 1.0.0
11	*/
12
13	namespace RadiusTheme\SB\AI\Embedding;
14
15	use RadiusTheme\SB\AI\DB\AIDB;
16	use RadiusTheme\SB\AI\AIFns;
17	use RadiusTheme\SB\Traits\SingletonTrait;
18
19	if ( ! defined( 'ABSPATH' ) ) {
20	exit( 'This script cannot be accessed directly.' );
21	}
22
23	/**
24	* Class DataEmbedding
25	*
26	* @package RadiusTheme\SB\AI\Embedding
27	* @since 1.0.0
28	*/
29	class DataEmbedding {
30	/**
31	* Singleton Trait.
32	*/
33	use SingletonTrait;
34
35	/**
36	* Generate and store an AI embedding for a product.
37	*
38	* @param int $product_id The product ID.
39	* @param string $title The product title.
40	* @param string $content The product content or description.
41	*
42	* @return bool True on success, false on failure.
43	*/
44	public function generate_and_store( $product_id, $title, $content ) {
45	$ai_data = AIFns::activated_ai_data();
46	if ( empty( $ai_data['api_key'] ) ) {
47	return false;
48	}
49
50	$text = $title . ' ' . wp_strip_all_tags( $content );
51	$ai_service = AIFns::initializeAIService();
52	$embedding = $ai_service->generateEmbedding( $text );
53	if ( empty( $embedding ) \|\| ! is_array( $embedding ) ) {
54	return false;
55	}
56	$info = [
57	'word_count' => str_word_count( $text ),
58	'source' => 'product',
59	];
60	$result = AIDB::upsert_embeding( $product_id, $title, $embedding, $info );
61	return ! empty( $result );
62	}
63
64	/**
65	* Perform a semantic search based on a given query.
66	*
67	* @param string $query The user search query.
68	* @param int $limit Optional. Number of results to return. Default 0 (all).
69	*
70	* @return array List of matching product titles.
71	*/
72	public function search( $query, $limit = 5 ) {
73	$ai_service = AIFns::initializeAIService();
74	$query_embedding = $ai_service->generateEmbedding( $query );
75	if ( empty( $query_embedding ) \|\| ! is_array( $query_embedding ) ) {
76	return [];
77	}
78	// Log query embedding.
79	$rows = AIDB::get_all();
80
81	if ( empty( $rows ) ) {
82	return [];
83	}
84	$results = $this->find_similar( $query_embedding, $rows, $limit );
85	return wp_list_pluck( $results, 'post_id' );
86	}
87
88	/**
89	* Find the most semantically similar embeddings using cosine similarity.
90	*
91	* Optimized to reduce redundant computations and use a dedicated
92	* score calculation method for better maintainability.
93	*
94	* @param array $embedding The query embedding vector to compare.
95	* @param array $rows The stored embedding records from the database.
96	* @param int $limit Optional. Number of top matches to return. Default 5.
97	*
98	* @return array List of matched items with product ID, title, and similarity score.
99	*/
100	public function find_similar( array $embedding, array $rows, int $limit = 5 ): array {
101	$minimum_match = AIFns::get_embedding_minimum_accuracy();
102	$query_norm = sqrt( array_sum( array_map( static fn( $x ) => $x * $x, $embedding ) ) );
103	$matches = [];
104	foreach ( $rows as $row ) {
105	if ( empty( $row['embedding'] ) ) {
106	continue;
107	}
108	$vector = maybe_unserialize( $row['embedding'] );
109	if ( ! is_array( $vector ) ) {
110	continue;
111	}
112	// Avoid redundant norm computation if similarity calc includes it.
113	$score = $this->calculate_similarity_score( $embedding, $vector, $query_norm );
114	if ( $score >= $minimum_match ) {
115	$matches[] = [
116	'post_id' => isset( $row['product_id'] ) ? (int) $row['product_id'] : 0,
117	'post_title' => isset( $row['title'] ) ? sanitize_text_field( $row['title'] ) : '',
118	'score' => round( $score, 4 ),
119	];
120	}
121	}
122	if ( empty( $matches ) ) {
123	return [];
124	}
125	// Use array_multisort for faster sorting on large datasets.
126	array_multisort( array_column( $matches, 'score' ), SORT_DESC, $matches );
127	return $limit > 0 ? array_slice( $matches, 0, $limit ) : $matches;
128	}
129
130
131	/**
132	* Calculate cosine similarity score between query and stored embedding.
133	*
134	* This version avoids redundant normalization by reusing the precomputed
135	* query norm and computes the dot product and target norm in one pass.
136	*
137	* @param array $query_vec The query embedding vector.
138	* @param array $target_vec The stored embedding vector.
139	* @param float $query_norm Precomputed L2 norm of the query vector.
140	*
141	* @return float Cosine similarity score (0.0–1.0).
142	*/
143	protected function calculate_similarity_score( array $query_vec, array $target_vec, float $query_norm ): float {
144	$dot = 0.0;
145	$normB = 0.0;
146	$len = min( count( $query_vec ), count( $target_vec ) );
147	for ( $i = 0; $i < $len; $i++ ) {
148	$dot += $query_vec[ $i ] * $target_vec[ $i ];
149	$normB += $target_vec[ $i ] ** 2;
150	}
151	if ( $query_norm <= 0.0 \|\| $normB <= 0.0 ) {
152	return 0.0;
153	}
154	return $dot / ( $query_norm * sqrt( $normB ) );
155	}
156	}
157