{"id":4473,"date":"2024-12-08T16:43:11","date_gmt":"2024-12-08T09:43:11","guid":{"rendered":"https:\/\/technica.ai\/?p=4473"},"modified":"2025-03-23T15:06:04","modified_gmt":"2025-03-23T08:06:04","slug":"knowledge-graph","status":"publish","type":"post","link":"https:\/\/technica.ai\/ja\/solutions\/knowledge-graph\/","title":{"rendered":"\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u3092\u6d3b\u7528\u3057\u305f\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9RAG"},"content":{"rendered":"<div data-elementor-type=\"wp-post\" data-elementor-id=\"4473\" class=\"elementor elementor-4473\">\n\t\t\t\t\t\t<section class=\"elementor-section elementor-top-section elementor-element elementor-element-9666658 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"9666658\" data-element_type=\"section\">\n\t\t\t\t\t\t<div class=\"elementor-container elementor-column-gap-default\">\n\t\t\t\t\t<div class=\"elementor-column elementor-col-100 elementor-top-column elementor-element elementor-element-6b18a11a\" data-id=\"6b18a11a\" data-element_type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t<div class=\"elementor-element elementor-element-9d2440c elementor-widget elementor-widget-text-editor\" data-id=\"9d2440c\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t<p>\u691c\u7d22\u62e1\u5f35\u751f\u6210\uff08RAG\uff09\u306f\u3001\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb\uff08LLM\uff09\u3068\u5916\u90e8\u77e5\u8b58\u30bd\u30fc\u30b9\u3092\u7d44\u307f\u5408\u308f\u305b\u308b\u3053\u3068\u3067\u3001\u6b63\u78ba\u3067\u95a2\u9023\u6027\u304c\u9ad8\u304f\u3001\u6587\u8108\u3092\u610f\u8b58\u3057\u305fAI\u30ec\u30b9\u30dd\u30f3\u30b9\u3092\u751f\u6210\u3057\u307e\u3059\u3002\u30d9\u30af\u30c8\u30eb\u30c7\u30fc\u30bf\u30d9\u30fc\u30b9\u306f\u5f93\u6765\u306eRAG\u30b7\u30b9\u30c6\u30e0\u306e\u4e2d\u5fc3\u7684\u306a\u5f79\u5272\u3092\u679c\u305f\u3057\u3066\u3044\u307e\u3059\u304c\u3001\u56fa\u6709\u306e\u5236\u9650\u304c\u3042\u308a\u307e\u3059\u3002\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u3092\u7d71\u5408\u3059\u308b\u3053\u3068\u3067\u3001RAG\u30b7\u30b9\u30c6\u30e0\u306f\u512a\u308c\u305f\u7cbe\u5ea6\u3001\u6587\u8108\u8a8d\u8b58\u3001\u8aac\u660e\u53ef\u80fd\u6027\u3092\u5b9f\u73fe\u3057\u3001AI\u30a2\u30d7\u30ea\u30b1\u30fc\u30b7\u30e7\u30f3\u306e\u65b0\u3057\u3044\u53ef\u80fd\u6027\u3092\u5207\u308a\u958b\u304d\u307e\u3059\u3002<\/p><h5>\u00a0<\/h5><h4>\u30d9\u30af\u30c8\u30ebDB\u306e\u307f\u306eRAG\u306b\u304a\u3051\u308b\u8ab2\u984c<\/h4><p>\u30d9\u30af\u30c8\u30eb\u30c7\u30fc\u30bf\u30d9\u30fc\u30b9\u306f\u610f\u5473\u7684\u985e\u4f3c\u6027\u306e\u6355\u6349\u306b\u512a\u308c\u3066\u3044\u307e\u3059\u304c\u3001\u91cd\u8981\u306a\u8ab2\u984c\u306b\u76f4\u9762\u3057\u3066\u3044\u307e\u3059\u3002<\/p><ol><li class=\"translation-block\">\u8907\u96d1\u306a\u95a2\u4fc2\u6027\uff1a\u30a8\u30f3\u30c6\u30a3\u30c6\u30a3\u3068\u6982\u5ff5\u9593\u306e\u8907\u96d1\u306a\u95a2\u4fc2\u6027\u306f\u3001\u30d9\u30af\u30c8\u30eb\u57cb\u3081\u8fbc\u307f\u3067\u306f\u5931\u308f\u308c\u304c\u3061\u3067\u3059\u3002<\/li><li class=\"translation-block\">\u66d6\u6627\u3055\uff1a\u985e\u4f3c\u3057\u305f\u57cb\u3081\u8fbc\u307f\u306b\u3088\u308a\u591a\u7fa9\u8a9e\u304c\u8aa4\u3063\u3066\u89e3\u91c8\u3055\u308c\u3001\u4e0d\u9069\u5207\u306a\u7d50\u679c\u306b\u3064\u306a\u304c\u308b\u53ef\u80fd\u6027\u304c\u3042\u308a\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u8aac\u660e\u53ef\u80fd\u6027\u306e\u6b20\u5982\uff1a\u30d9\u30af\u30c8\u30eb\u985e\u4f3c\u6027\u306e\u307f\u306b\u57fa\u3065\u304f\u691c\u7d22\u5224\u65ad\u306f\u4e0d\u900f\u660e\u3067\u89e3\u91c8\u304c\u56f0\u96e3\u3067\u3059\u3002<\/li><\/ol><h5>\u00a0<\/h5><h4>\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u3092\u6d3b\u7528\u3057\u305f\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9RAG<\/h4><p>\u30d9\u30af\u30c8\u30eb\u30c7\u30fc\u30bf\u30d9\u30fc\u30b9\u3068\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u306e\u9577\u6240\u3092\u7d71\u5408\u3057\u305f\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9RAG\u30a2\u30fc\u30ad\u30c6\u30af\u30c1\u30e3\u3092\u63d0\u6848\u3057\u307e\u3059\u3002<\/p><ol><li class=\"translation-block\">\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u69cb\u7bc9\uff1a\u30c7\u30fc\u30bf\u304b\u3089\u30a8\u30f3\u30c6\u30a3\u30c6\u30a3\u3001\u6982\u5ff5\u3001\u305d\u308c\u3089\u306e\u76f8\u4e92\u95a2\u4fc2\u306e\u69cb\u9020\u5316\u3055\u308c\u305f\u8868\u73fe\u3092\u69cb\u7bc9\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u30c7\u30e5\u30a2\u30eb\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\uff1a\u30c7\u30fc\u30bf\u306f\u30d9\u30af\u30c8\u30eb\u30c7\u30fc\u30bf\u30d9\u30fc\u30b9\u3068\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u306e\u4e21\u65b9\u306b\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u5316\u3055\u308c\u3001\u76f8\u88dc\u7684\u306a\u691c\u7d22\u30e1\u30ab\u30cb\u30ba\u30e0\u3092\u53ef\u80fd\u306b\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u30af\u30a8\u30ea\u5f37\u5316\uff1a\u30e6\u30fc\u30b6\u30fc\u30af\u30a8\u30ea\u304b\u3089\u91cd\u8981\u306a\u30a8\u30f3\u30c6\u30a3\u30c6\u30a3\u3092\u62bd\u51fa\u3057\u3001\u305d\u308c\u3089\u306e\u95a2\u4fc2\u6027\u3092\u5206\u6790\u3057\u3001\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u304b\u3089\u95a2\u9023\u7528\u8a9e\u3084\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u3092\u7528\u3044\u3066\u30af\u30a8\u30ea\u3092\u62e1\u5f35\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9\u691c\u7d22\uff1a\u5f37\u5316\u3055\u308c\u305f\u30af\u30a8\u30ea\u306b\u3088\u308a\u3001\u30d9\u30af\u30c8\u30eb\u30c7\u30fc\u30bf\u30d9\u30fc\u30b9\u3068\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u306e\u4e21\u65b9\u304b\u3089\u76f8\u88dc\u7684\u306a\u60c5\u5831\u3092\u53d6\u5f97\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u30ec\u30b9\u30dd\u30f3\u30b9\u751f\u6210\uff1a\u7d71\u5408\u3055\u308c\u305f\u30c7\u30fc\u30bf\u306fLLM\u3092\u4f7f\u7528\u3057\u3066\u3001\u6b63\u78ba\u3067\u6587\u8108\u306b\u95a2\u9023\u3057\u305f\u5fdc\u7b54\u306b\u5408\u6210\u3055\u308c\u307e\u3059\u3002<\/li><\/ol><h5>\u00a0<\/h5><h4><strong>\u30e1\u30ea\u30c3\u30c8<\/strong><\/h4><ol><li class=\"translation-block\">\u7cbe\u5ea6\u306e\u5411\u4e0a\uff1a\u660e\u78ba\u306b\u5b9a\u7fa9\u3055\u308c\u305f\u30a8\u30f3\u30c6\u30a3\u30c6\u30a3\u3068\u95a2\u4fc2\u6027\u306b\u691c\u7d22\u3092\u56fa\u5b9a\u3059\u308b\u3053\u3068\u3067\u3001\u66d6\u6627\u3055\u3092\u89e3\u6d88\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u95a2\u9023\u6027\u306e\u5411\u4e0a\uff1a\u6587\u8108\u3092\u610f\u8b58\u3057\u305f\u691c\u7d22\u306b\u3088\u308a\u3001\u30e6\u30fc\u30b6\u30fc\u306e\u610f\u56f3\u306b\u5bc6\u63a5\u306b\u6cbf\u3063\u305f\u7d50\u679c\u3092\u78ba\u4fdd\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u8aac\u660e\u53ef\u80fd\u6027\uff1a\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u306e\u69cb\u9020\u5316\u3055\u308c\u305f\u6027\u8cea\u306b\u3088\u308a\u3001\u7279\u5b9a\u306e\u7d50\u679c\u304c\u8fd4\u3055\u308c\u308b\u7406\u7531\u306e\u900f\u660e\u6027\u304c\u63d0\u4f9b\u3055\u308c\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u8907\u96d1\u306a\u63a8\u8ad6\uff1a\u3088\u308a\u6df1\u3044\u63a8\u8ad6\u3068\u63a8\u6e2c\u3092\u53ef\u80fd\u306b\u3057\u3001\u9ad8\u5ea6\u306a\u554f\u984c\u89e3\u6c7a\u80fd\u529b\u3092\u5b9f\u73fe\u3057\u307e\u3059\u3002<\/li><\/ol><h5>\u00a0<\/h5><h4><strong>\u30e6\u30fc\u30b9\u30b1\u30fc\u30b9<\/strong><\/h4><p>\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9RAG\u30a2\u30d7\u30ed\u30fc\u30c1\u306f\u3001\u8907\u6570\u306e\u5206\u91ce\u3067\u9769\u65b0\u7684\u306a\u53ef\u80fd\u6027\u3092\u63d0\u4f9b\u3057\u307e\u3059\u3002<\/p><ul><li class=\"translation-block\">\u30a4\u30f3\u30c6\u30ea\u30b8\u30a7\u30f3\u30c8\u691c\u7d22\uff1a\u30ad\u30fc\u30ef\u30fc\u30c9\u30de\u30c3\u30c1\u30f3\u30b0\u3092\u8d85\u3048\u305f\u3001\u6b63\u78ba\u3067\u6587\u8108\u3092\u610f\u8b58\u3057\u305f\u691c\u7d22\u7d50\u679c\u3092\u30e6\u30fc\u30b6\u30fc\u306b\u63d0\u4f9b\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u5bfe\u8a71\u578b\u30c1\u30e3\u30c3\u30c8\u30dc\u30c3\u30c8\uff1a\u9ad8\u5ea6\u306b\u9b45\u529b\u7684\u3067\u77e5\u8b58\u8c4a\u5bcc\u306a\u5bfe\u8a71\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/li><li class=\"translation-block\">\u30ec\u30b3\u30e1\u30f3\u30c7\u30fc\u30b7\u30e7\u30f3\u30b7\u30b9\u30c6\u30e0\uff1a\u30e6\u30fc\u30b6\u30fc\u306e\u5fae\u5999\u306a\u597d\u307f\u306b\u57fa\u3065\u3044\u305f\u3001\u30d1\u30fc\u30bd\u30ca\u30e9\u30a4\u30ba\u3055\u308c\u305f\u6587\u8108\u306b\u95a2\u9023\u3059\u308b\u30ec\u30b3\u30e1\u30f3\u30c7\u30fc\u30b7\u30e7\u30f3\u3092\u63d0\u4f9b\u3057\u307e\u3059\u3002<\/li><\/ul><p>\u00a0<\/p><p>\u30ca\u30ec\u30c3\u30b8\u30b0\u30e9\u30d5\u3092\u6d3b\u7528\u3057\u305f\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9RAG\u3067\u3042\u306a\u305f\u306eAI\u30b7\u30b9\u30c6\u30e0\u306e\u53ef\u80fd\u6027\u3092\u6700\u5927\u9650\u306b\u5f15\u304d\u51fa\u3057\u307e\u3057\u3087\u3046\u3002\u9769\u65b0\u7684\u306a\u30bd\u30ea\u30e5\u30fc\u30b7\u30e7\u30f3\u304c\u3069\u306e\u3088\u3046\u306b\u60c5\u5831\u691c\u7d22\u3068\u751f\u6210\u6a5f\u80fd\u3092\u9769\u65b0\u3067\u304d\u308b\u304b\u3001\u4eca\u3059\u3050\u3054\u76f8\u8ac7\u304f\u3060\u3055\u3044\u3002<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<\/div>","protected":false},"excerpt":{"rendered":"<p>Retrieval Augmented Generation (RAG) combines the power of large language models (LLMs) with external knowledge sources to produce accurate, relevant, and context-aware AI responses. While vector databases are central to traditional RAG systems, they have inherent limitations. By integrating knowledge graphs, RAG systems achieve superior precision, context awareness, and explainability, unlocking new possibilities for AI applications. \u00a0 Challenges with Vector DB-only RAG Vector databases excel at capturing semantic similarity but face significant challenges: Complex Relationships: Intricate relationships between entities and concepts are often lost in vector embeddings. Ambiguity: Similar embeddings can misinterpret polysemous terms, leading to irrelevant results. Lack of Explainability: Retrieval decisions based solely on vector similarity are opaque [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":5608,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"footnotes":""},"categories":[13],"tags":[58,57],"class_list":["post-4473","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-solutions","tag-hybrid-rag","tag-knowledge-graph"],"acf":[],"_links":{"self":[{"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/posts\/4473","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/comments?post=4473"}],"version-history":[{"count":1,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/posts\/4473\/revisions"}],"predecessor-version":[{"id":5609,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/posts\/4473\/revisions\/5609"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/media\/5608"}],"wp:attachment":[{"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/media?parent=4473"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/categories?post=4473"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/technica.ai\/ja\/wp-json\/wp\/v2\/tags?post=4473"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}