KNowledge Acquisition and Representation Methodology (KNARM)
KNowledge Acquisition and Representation Methodology (KNARM) allows domain experts and knowledge engineers to build useful, consistent, concordant knowledge graphs and ontologies formalizing domain data and knowledge in a systematic way using modular ontology architecture and systematically deepening modeling for domain knowledge. It is designed to help with the challenge of acquiring and representing knowledge in a systematic, semi-automated way. This methodology aims at acquiring knowledge from data scattered in different databases and ontologies, combining them in a meaningful fashion that is understandable by humans and machines by effectively combining human and machine capabilities. In this way, we attempt to allow users to understand, query, and analyze their data better.
In this session, we will go over the steps of creating a modular ontology using KNARM which serves the PFAS Use Cases. At each step of the methodology, we will discuss what was done, and how it could be improved using systematically deepening modeling and using a systematic approach to building the knowledge graph backend with a semi-automated approach using a modular architecture.
- Sub-language Analysis
This is already done for us: PFAS Use Case.
- Unstructured Interview
This step combined with the previous step is all about understanding the questions better.
- Sub-Language Recycling
Identify vocabularies and patterns that may have already been generated to avoid duplication of efforts.
- Meta Data Creation and Knowledge Modeling
Based on the questions and sub-language identified, can we create our modeling of the data? Questions to consider: how deep do I go in my modeling? What aspects are modeled in T-Box, and what’s modeled in A-Box?
- Structured Interview
Use the model to retrieve data from domain experts in a FAIR way from the start.
- Knowledge Acquisition Validation
Ensure that we can still answer the complex questions we had at the beginning..
- Database Formation
Opt for sustainable data storage and create a database backend.
- Semi-Automated Ontology Building
Use tools like Robot or write your own scripts for building the owl files from the database backend following a modular architecture and systematically deepening modeling outlined in KNARM.
- Ontology Validation
Ultimate test for your ontology is to put it in use via applications, community, and algorithms and evaluate performance. Keep in mind that KNARM follows an agile approach and believes in improving the knowledge graphs based on community needs and changes to the vocabularies in the backend.
References & Resources
Here are some convenient links to tools, resources, and examples to use while implementing your knowledge graphs using KNARM.
Ontologies Built using KNARM
- Protégé – for file generation and manual steps
- Robot – a command-line tool for creating .owl files in a semi-automated way.
- Neo4j – for the graph database backend