Elasticsearch2.4.4自定义词典&同义词配置

自定义词典：

一、添加词典

mkdir -p elasticsearch-2.4.4/plugins/analysis-ik/config/custom

vi elasticsearch-2.4.4/plugins/analysis-ik/config/custom/ext_word.txt

博世
bosch

注意事项：

1，每个单词一行

2，编码为utf-8 无bom

二、修改ik配置

"1.0" encoding="UTF-8"?> "http://java.sun.com/dtd/properties.dtd">     IK Analyzer 扩展配置          "ext_dict">custom/ext_word.dic;custom/single_word_low_freq.dic           "ext_stopwords">custom/ext_stopword.dic

三、重启es

同义词配置：

一、添加词典：

mkdir -p elasticsearch-2.4.4/config/analysis

vi elasticsearch-2.4.4/config/analysis/synonym.txt

博世,bosch

注意事项：

1，每行一组同义词，以逗号分隔

2，编码为utf-8 无bom

3，修改synonym.txt后需要重启es

二、索引配置修改

新建业务索引2_syn，添加同义词过滤器synonym_filter

setting设置如下：

{   "index": {     "analysis": {       "filter": {         "light_english_stemmer": {           "type": "stemmer",           "language": "light_english"         },         "special_character_spliter": {           "type": "word_delimiter",           "preserve_original": "true"         },         "synonym_filter": {           "type": "synonym",           "synonyms_path": "analysis/synonym.txt"         }       },       "analyzer": {         "charSplit": {           "filter": [             "lowercase",             "synonym_filter"           ],           "type": "custom",           "tokenizer": "ngram_tokenizer"         },         "optik_smart": {           "filter": [             "lowercase",             "light_english_stemmer",             "special_character_spliter",             "synonym_filter"           ],           "type": "custom",           "tokenizer": "ik_smart"         },         "optik": {           "filter": [             "lowercase",             "light_english_stemmer",             "special_character_spliter",             "synonym_filter"           ],           "type": "custom",           "tokenizer": "ik"         }       },       "tokenizer": {         "ngram_tokenizer": {           "token_chars": [             "letter",             "digit",             "punctuation"           ],           "min_gram": "1",           "type": "nGram",           "max_gram": "30"         }       }     }   } }

 

三、测试同义词

GET   /2_syn/_analyze?analyzer=optik&pretty=true&text=博世

结果：

{   "tokens": [     {       "token": "博世",       "start_offset": 0,       "end_offset": 2,       "type": "CN_WORD",       "position": 0     },     {       "token": "bosch",       "start_offset": 0,       "end_offset": 2,       "type": "SYNONYM",       "position": 0     }   ] }

 

四、数据迁移

使用reindex api迁移数据

POST _reindex {   "source": {     "index": "2"   },   "dest": {     "index": "2_syn"   } }

 

问题：

1，修改同义词词典synonym.txt 需要重启es

2，ik无法正确分词的token无法找到同义词，需要配合自定义词库使用