Knowledge Query Language (KQL)
Abstract
Currently, queries for retrieval from NoSQL data stores are tightly coupled to the specific implementation of the data store implementation, making portability of the queries or query-dependent algorithms difficult. This report introduces a declarative approach that is independent of the storage content and format for querying NoSQL or relational data stores. This approach uses address expressions (or A-Expressions) embedded in commonly used query languages such as Structured Query Language (SQL). The declarative approach makes the queries portable, and results in several advantages over the existing approaches to querying, especially when the data is semi-structured, and when the data sources may change over time. Cyber event logs are examples of such data sources. When the query is independent of the underlying physical data sources, having provenance information on the query results becomes important to impart necessary context, and ensure trust in the query results returned. This declarative approach is made possible through the use of a Knowledge Registry. In this report, we discuss embedding A-Expressions in the widely used SQL, resolving A-Expressions using the ontology implemented in a Knowledge Registry, and returning query results with provenance information.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 2016
- Accession Number
- AD1011590
Entities
People
- Suresh K. Damodaran
Organizations
- Massachusetts Institute of Technology