New Feature, CLO-SET Elastic Search Engine!

Feature released: 2022-10-18

Hello CLO-SET Users, 

CLO-SET Elastic Search Engine is newly deployed.

This is mostly our engine update on our back-end, but there are slight enhancements for your user experience as well.

Benefits

• Faster Query
• More Accurate Search Results
• Wildcard Query
• Category keyword Search

What Is Elasticsearch?

Some of you may already be familiar with Elasticsearch, and others may not be. So let us briefly explain just some of the key points.

Elasticsearch is a distributed, free, and open search and analytics engine for all types of data. It is widely used for its advantages in search and analytics due to its ability to index different data types. Elasticsearch has a process called data ingestion, in which the raw data is parsed, normalized, and enriched, and then indexed. 

An Elasticsearch index is a collection of documents - JSON documents where data is stored. And, Elasticsearch uses a data structure called an inverted index, which allows very quick full-text searches. 

If you want to know more about Elasticsearch in depth, please check out https://www.elastic.co/what-is/elasticsearch.

Enhancements

Before, when users searched in CLO-SET, the SQL query was performed on our SQL database by looking up the right tables and finding the search results.

But now, with Elasticsearch, when users search in CLO-SET, the search will be performed by calling APIs in the Index created in Elasticsearch.

• Faster Query
  • Elasticsearch is near real-time because Elasticsearch has an indexing process. It builds data in JSON documents, and builds inverted index making the document data search done in near real-time instead of having to join tables in the database to find the search results as previously done. 
    
• More Accurate Search Results
  • The search results are sorted by their scores that are computed by our specific computation method. 
    • For example, when searched for ‘zpac sample’, the results with ‘zpac’ are given 20 points and ‘sample’ is given 10 points. 
    • Then, among those results, priorities are also sorted.
    • The results would look like the below.
      • Original SQL query will not show any results as no file is actually named “zpac sample”.

• Wildcard Query
  • A wildcard operator is a placeholder that matches one or more characters.

Roadmap for CLO-SET

Now that the search engine is implemented in CLO-SET, we will work towards the UX/UI improvements in search feature that will strengthen and optimize the user experience. We will keep you posted!