전문 검색(Full-Text Search) 색인 정의를 위한 스키마

{
    "name": "sk_fts",
    "fields": [
        "doc_id",
        "content",
        "embedding"
    ],
    "unique": False,
    "index_type": "kFullTextSearchIndex",
    "options": {
        "doc_id": {
            "analyzer": {
                "type": "KeywordAnalyzer"
            },
            "index_options": "doc_freqs"
        },
        "content": {
          "analyzer": {
            "type": "StandardAnalyzer"
          },
          "index_options": "offsets"
        },
        "embedding": {
            "analyzer": {
                "type": "DenseVectorAnalyzer",
                "options": {
                    "index_type": "HNSW",
                    "dims": 768,
                    "m": 64,
                    "ef_construction": 200,
                    "ef_search": 32,
                    "metric": "inner_product",
                    "normalize": True,
                    "shards": 1
                }
            },
            "index_options": "doc_freqs"
        }
    }
}

{
    "doc_id": {
        "index_options": "doc_freqs",
        "analyzer": {
            "type": "KeywordAnalyzer"
        }
    }
},

{
    "type": "KeywordAnalyzer"
}

{
    "type": "keyword"
}

{
    "type": "CustomAnalyzer",
    "options": {
        "char_filters": [
            "LowerCaseCharacterFilter"
        ],
        "tokenizer": "ICUWordTokenizer",
        "token_filters": [
            "NormalizationTokenFilter",
            {
                "type": "SnowballTokenFilter",
                "options": {
                    "language": "english"
                }
            }
        ]
    }
}

{
    "type": "KeywordAnalyzer"
    "options": {
        "case_insensitive": True
    }
}

{
    "type": "CustomAnalyzer"
    "options": {
        "char_filters": [
            "LowerCaseCharacterFilter", "NormalizationCharacterFilter"
        ],
        "tokenizer": "ICUWordTokenizer",
        "token_filters": [
            {
                "type": "StopWordsTokenFilter",
                "options": {
                    "language": "english"
                }
            },
            {
                "type": "SnowballTokenFilter",
                "options": {
                    "language": "english"
                }
            }
        ]
    }
}

{
    "options": {
        "stopwords_filter": {
          "enabled": True
        },
        "snowball_filter": {
          "enabled": True
        }
    }
}

{
    "type": "StandardCJKAnalyzer"
    "options": {
        "ngram_filter": {
            "min_size": 1,
            "max_size": 6
        },
        "byte_length_filter": {
            "min_length": 2,
            "max_length": 0
        }
    }
}

{
    "type": "DenseVectorAnalyzer",
    "options": {
        "index_type": "HNSW",
        "dims": 768,
        "m": 64,
        "ef_construction": 200,
        "ef_search": 32,
        "metric": "inner_product",
        "normalize": True,
        "shards": 1
    }
}

{
    "type": "RegexAnalyzer",
    "options": {
        "tokenizer": {
            "type": "RegexTokenizer",
            "options": {
                "patterns": [
                    "(\\b|\\$)\\d+(,\\d+)*[\\-\\w%]*",
                    "[a-zA-Zㄱ-ㅎ가-힣][&0-9a-zA-Zㄱ-ㅎ가-힣]*"
                ],
                "trim": ",- "
            }
        }
    }
}

{
    "type": "DelimiterTokenizer",
    "options": {
        "delimiter": ", "
    }
}

Aeca API, Documents

["Aeca", "API", "Documents"]

{
    "type": "NGramTokenizer",
    "options": {
        "min_size": 2,
        "max_size": 4
    }
}

코그니카

["코그", "그니", "니카", "코그니", "그니카", "코그니카"]

{
    "type": "RegexTokenizer",
    "options": {
        "patterns": [
            "(\\b|\\$)\\d+(,\\d+)*[\\-\\w%]*",
            "[a-zA-Zㄱ-ㅎ가-힣][&0-9a-zA-Zㄱ-ㅎ가-힣]*"
        ],
        "trim": ",- "
    }
}

{
    "type": "PathHierarchyTokenizer",
    "options": {
        "delimiter": "/"
    }
}

a/b/c

["a", "a/b", "a/b/c"]

{
    "type": "ReversePathHierarchyTokenizer",
    "options": {
        "delimiter": "/"
    }
}

a/b/c

["a/b/c", "a/b", "a"]

NFD; [:Mn:] Remove; NFC

NFD; [:Mn:] Remove; NFC

{
    "type": "CharLengthTokenFilter",
    "options": {
        "min_length": 2,
        "max_length": 4,
    }
}

["다람쥐", "헌", "쳇바퀴에", "타고파"]

["다람쥐", "쳇바퀴에", "타고파"]

{
    "type": "NGramTokenFilter",
    "options": {
        "min_size": 2,
        "max_size": 4,
    }
}

["코그니카"]

["코그", "그니", "니카", "코그니", "그니카", "코그니카"]

{
    "type": "EdgeNGramTokenFilter",
    "options": {
        "min_size": "2",
        "max_size": "3",
    }
}

["코그니카", "데이터베이스는"]

["코그", "코그니", "데이", "데이터"]

{
    "type": "StopWordsTokenFilter",
    "options": {
        "language": "english"
    }
}

{
    "type": "ShingleTokenFilter",
    "options": {
        "min_size": 2,
        "max_size": 3,
        "output_unigrams": True,
        "seperator": " ",
    }
}

["애플", "컴퓨터", "신규", "제품"]

[
    "애플", "애플 컴퓨터", "애플 컴퓨터 신규",
    "컴퓨터", "컴퓨터 신규", "컴퓨터 신규 제품",
    "신규", "신규 제품",
    "제품"
]

{
    "type": "SnowballTokenFilter",
    "options": {
        "language": "english"
    }
}