How to use Elasticsearch to implement pre input search

What is pre input search?

Pre input search, also known as auto suggestion or auto completion, is a method of filtering data by checking whether the user input data is a subset of data. If so, all text that partially matches the user is a way to provide a prompt when typing text. This feature undoubtedly contributes to the end user's search experience.

How to use Elasticsearch to realize pre input search

Elastic search provides four different ways to implement pre input search. Let's look at all four methods to see which is the best and better implemented:

• Match phrase prefix
• Edge Ngram
• Complete recommendations
• Search on input

Match phrase prefix query

In this approach, we need to use prefix queries on the search fields. The query returns documents containing the words of the prefix text provided in the same order as provided.  

For example:

GET /_search { "query": { "match_phrase_prefix": { "message": { "query": "quick brown f" } } } }

The above search returns the document quick brown fox or two quick brown ferrets, but not fox is quick and brown, because the query will match the phrase beginning with quick brown f -

Edge Ngram

Using this approach, we need to configure the custom analyzer using the edge-n-gram filter.

PUT myIndex { "settings": { "analysis": { "analyzer": { "autocomplete": { "tokenizer": "autocomplete", "filter": [ "lowercase" ] }, "autocomplete_search": { "tokenizer": "lowercase" } }, "tokenizer": { "autocomplete": { "type": "edge_ngram", "min_gram": 2, "max_gram": 5, "token_chars": [ "letter" ] } } } }, "mappings": { "properties": { "title": { "type": "text", "analyzer": "autocomplete", "search_analyzer": "autocomplete_search" } } } }

Several points to consider when implementing this method:

• It is best to use the same parser for index and search
• Indexing time may increase because the marker breaks text into words on custom characters
• As mentioned above, index storage will be increased

Complete recommendations

Elasticsearch provides a Completion Suggester as a native solution to the auto complete / search as input function. Autocomplete should be as fast as user input to provide immediate feedback on what the user has entered. Therefore, the completion recommender uses data structures that support fast lookup but are expensive to build and stored in memory.

How do I set the completion advisor?

To use Completion Suggester, we need to make a special mapping for this field:

PUT movie { "mappings": { "_doc" : { "properties" : { "suggest" : { "type" : "completion" }, "title" : { "type": "keyword" } } } } }

Index data:

PUT movie/_doc/1?refresh { "suggest" : [ "Elevate Me Later", "Covert affairs" ] }

Ask:

POST movie/_search?pretty { "suggest": { "movie-suggest" : { "prefix" : "ele", "completion" : { "field" : "suggest", "skip_duplicates": true } } } }

Search on input

Search as you type is an optimized field type that provides out of the box support for pre entering search or as you type completion of use cases. The search as you type map creates many subfields and indexes the data by analyzing terms, which helps to partially match the indexed text values. It supports prefix completion and infix completion.

To configure search as you type, add the following mappings to your index fields:

PUT search-index { "mappings": { "properties": { "title": { "type": "search_as_you_type" } } } }

You can view the search as you type or complete the suggestion program to use Elasticsearch for pre input. Although using matching phrase prefix or edge-n-gram, we can achieve the same effect, that is, search by type, wrap this implementation internally, and provide efficient and optimized functions as field types.

Pepper shell - may you be safe and healthy. Pepper shell resource network, resource download, mainly provides personal collection resources, design materials, music, video, pictures and other Internet related resourceshttps://www.xinliu.vip/

Add new comment

0 comments