File size: 3,125 Bytes
b321188
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6bdb67d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pymilvus import MilvusClient\n",
    "\n",
    "\n",
    "client = MilvusClient(uri=\"./db/Amazon_electronics.db\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cd7440ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['image_embeddings', 'metadata_embeddings', 'rating_embeddings']"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "client.list_collections()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ef468e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def query_image_similarity(client, asin, top_k=50):\n",
    "    \"\"\"\n",
    "    查询指定 asin 对应的图片向量,并在 Milvus 中搜索相似商品(图片相似度)。\n",
    "\n",
    "    返回:\n",
    "      字典格式 {asin: sim_score}\n",
    "      其中 sim_score 采用 COSINE 指标,计算方式: sim_score = 1 - hit.distance\n",
    "    \"\"\"\n",
    "    try:\n",
    "        query_expr = f\"asin == '{asin}'\"\n",
    "        query_res = client.query(\n",
    "            collection_name=\"image_embeddings\",\n",
    "            filter=query_expr,\n",
    "            output_fields=[\"embedding\"],\n",
    "        )\n",
    "        if not query_res:\n",
    "            return {}\n",
    "\n",
    "        target_embedding = query_res[0][\"embedding\"]\n",
    "        search_params = {\"metric_type\": \"COSINE\", \"params\": {\"nprobe\": 10}}\n",
    "        search_results = client.search(\n",
    "            collection_name=\"image_embeddings\",\n",
    "            data=[target_embedding],\n",
    "            anns_field=\"embedding\",\n",
    "            search_params=search_params, \n",
    "            limit=top_k,\n",
    "            filter=f\"asin != '{asin}'\",  # 排除自身\n",
    "        )\n",
    "\n",
    "        sim_dict = {}\n",
    "        for hit in search_results[0]:\n",
    "            sim_asin = hit[\"id\"]\n",
    "            sim_score = 1 - hit[\"distance\"]\n",
    "            sim_dict[sim_asin] = sim_score\n",
    "        return sim_dict\n",
    "    except Exception as e:\n",
    "        print(f\"图片相似度查询失败: {e}\")\n",
    "        return {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3bec7aa1",
   "metadata": {},
   "outputs": [],
   "source": [
    "query_image_similarity(client, \"B07X2Y3Z5F\", top_k=10)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv (3.12.3)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}