Use Machine Learning Annotator Plugin to enable ML annotator in oneWEX

Create a Machine Learning Annotator Plugin enrichment for new collections

1. Build Machine Learning Annotator Plugin for Custom Enrichment

You will need to copy /opt/ibm/wex/nlp/lib in oneWEX to lib to resolve the dependencies.

• Edit plugins/mla-plugin/src/main/resources/server.properties, Configure WEX AC server

  wexac.host=localhost
  wexac.port=8393
  wexac.collectionId=ml
  

• run gradle build to build Mahchine Learning Annotator plugin, find the plugin archive under plugin folder.

  Note 1: Gradle 5 is not available. Use Gradle 4.

  Note 2: Implement only one implementation of CustomEnrichment

• Create file resource

• Upload Mahchine Learning Annotator Plugin

2. Configure CustomEnrichment in oneWEX

• Get facet mappings from WEX AC

  • Get security token
  • Get facet tree
  • Convert to oneWEX facet

• Create Enrichment

  • reate enrichment with file resource

• Enable Mahchine Learning Annotator enrichment

  • Enable the enrichment and rebuild the index

Import Tool

For Step 2, an import tool is provided for converting and importing the facet mappings

Go to importer and run ./import.sh with following environment variables. Bash and Python3 is required.

Name	Description
ONEWEX_HOST	oneWEX host
WEXAC_HOST	WEX AC host
WEXAC_USER	WEX AC user name, esadmin by default
WEXAC_PASSWORD	WEX AC user password, prompt for input if missing
WEXAC_COLLECTION_ID	WEX AC collection ID
PATH_TO_PLUGIN	Path to Mahchine Learning Annotator plugin, build/plugins/plugin-mla-plugin-0.0.1-SNAPSHOT.zip by default

• Example

ONEWEX_HOST=localhost WEXAC_HOST=localhost WEXAC_COLLECTION_ID=ml ./import.sh