maythapk / esninja

Best practices for product search in English and Thai using Elasticsearch

esninja

Best practices for product search in English and Thai using Elasticsearch

Getting Started

1. Install Elastic tack. For this tutorial we only need:

• ElasticSearch
• Kibana
• Logstash; may require JDK
• Install ICU tokenizer plugin

2. Run elasticsearch locally (default port is 9200), in elasticsearch folder:

./bin/elasticsearch

3. Import the sample .csv file into elasticsearch using logstash (see field data types for reference):

#do not forget to change path to files in tops_sample.conf
logstash -f tops_sample.conf

4. Run kibana for the local elasticsearch instance (default port is 5601), in kibana folder:

./bin/kibana

5. Run tops_sample_optimized.json on kibana to create new mappings with optimized analyzers.

6. Follow tops_sample_dev.json for how to configure mappings, settings and queries.

Docker

Still not available to public.

docker-compose up -d

Miscellaneous Ninja Tools

• elasticdump - Tools for moving and saving indices; requires npm.
• elasticsearch head - Chrome addon for index visualization