Main Genny Project

Genny (short for Generic) is an open source project that serves to create an intelligent "Question and Answer" system.

Table of Contents

1. Technologies
2. Components
3. Install
4. Genny Documentation
   • Buckets
   • Add Items

Technologies

The technologies used:

• Docker Swarm
• Java Wildfly Swarm
• Vertx
• React
• Keycloak
• Drools Rules/Workflow
• MySQL
• Quarkus

Intended technologies:

• Esper (CEP)
• Tensorflow (for scoring, Deep Learning)
• GraphQL
• Apache Spark (scoring)
• iOS, Android (React Native, Flutter)
• Neo4J,Cassandra

Components

The main components of the Genny project are:

Qwanda Engine - This section of the arcitecture manages the storage, querying, scoring of the Questions and Answers.

Bridge - This Vertx service acts as a message bridge between the Vertx Event Bus and the frontend Websocket channels.

Alyson - This is a React based generic frontend that processes cmds and data sent from the backend and displays the requested display and data. It has integrated Keycloak security and Vertx integration for messaging.

Social - This Vertx service manages external social media data querying.

Wildfly-RulesService - This Wildfly Rules engine service performs fast Esper/Drools Complex Event Processing (CEP) RulesService - This Vertx service performs fast Esper/Drools Complex Event Processing (CEP)

Install

Assume you have java 11 jdk and mvn 3.5+ installed along with git and docker and docker-compose

1. create a project folder called 'genny'

mkdir genny

2. enter the genny folder and pull the genny-main project from git

cd genny
git clone -b v3.1.0 https://github.com/genny-project/genny-main

3. Now fetch all the other projects

cd genny-main
./cyrusClone.sh

4. Now build them all

./cyrusBuild.sh

To run: (simple)

run the docker versions on your local machine -> ./run.sh <project> up

Genny Documentation

Buckets

To establish a bucket page in Genny, the following must be set up:

1. A set of SearchEntitys must be built and placed in the GenerateResources/GenerateSearches/Buckets rulegroup directory.

2. Inside the file GenerateResources/GenerateAsks/GENERATE_BUCKET_CODES.drl, you must provide a list for caching of these search codes in the order of buckets from left to right.

3. In INIT_GENERATE_CAPABILITIES.drl you must create a seperate capability code for each of these search codes and assign viewing capabilities to the required roles for each of these capability codes

Example in Role:

capabilityUtils.addCapabilityToRole(role, "SBE_AVAILABLE_INTERNS", CapabilityMode.VIEW);

Add Items

2. Inside the file GenerateResources/GenerateAsks/GENERATE_ADD_ITEMS.drl, you must provide a LinkedHashmap for caching that contains the role (Prefixed with "QA" if quck add) and the corresponding text to display.

3. In INIT_GENERATE_CAPABILITIES.drl there must exist a capability code for each type in the system (E.g TYPE_INTERN), and ADD capability must be provided for the required roles.

Example in Role:

capabilityUtils.addCapabilityToRole(role, "TYPE_INTERN", CapabilityMode.ADD);