VExSS is a prototype expert system designed to target a single vendor or open source project (OSP) from the Vendelligence directory of resources.
I built the original prototype in 2015 when I was playing with Elasticsearch 1.x releases on Windows 10. This is an updated version covering Spring projects instead of the SAP domain. It runs on Elasticsearch 5.x.
Kibana was used for indexing in the interim, as it lets me iterate the JSON data model quickly in tandem with my text editors of choice, without needing me to write a separate CRUD utility during the prototyping phase. I have since turned to developing a simple batch tool using Spring Batch and Spring Integration to prototype the utility.
I wanted a dedicated expert system to support me during my own vendor support services or prototype experiments with open source projects (OSPs). These tools are attempts to build it via trial-and-error essentially.
While Vendelligence focuses on the data curation and data storage steps, VExSS focuses on querying that stored Query data, using a more complex data structure, to help lay the building blocks for an initial expert system.
The idea is to tune the search results using the flags, labels, and pattern variants in the JSON data model instances.
The objective is to get the Elasticsearch simple version working first, create a data ingestion workflow via Forklift, and and then try out expert-system rules engine solutions to work with the data.
VExSS is updated on GitHub.