Open-Source Toolkit for Knowledge-Based Querying of Time-Oriented Data

Stanford University, Stanford CA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): In research involving biological pathways, disease progression, and treatment outcomes, investigators face particular difficulties in identifying important relationships among the temporal information they collect. Currently, investigators have limited abilities to discover relevant temporal patterns among volumes of observational and experimental data based, because of the lack of widely available, ready-to-use tools that incorporate temporal and domain knowledge. To address these challenges, we propose a novel method for querying and abstracting temporal patterns that integrates data, ontologies, rules, and formal temporal semantics. Our approach, called Semantic Query-enhanced Web Rule Language (SQWRL), is an extension to W3C standards for the Semantic Web and to the Protigi environment-the mostly widely used, freely available, open-source software for specifying ontologies and knowledge bases. The proposed querying toolkit, called SQWRL-TK, will provide investigators an integrated, scalable framework for domain-driven investigation of temporal phenomena. The goal of our efforts is to develop a set of open-source tools that flexibly integrates knowledge-based methods for the transformation, retrieval, and abstraction of temporal data. Our specific aims are (1) to implement a scalable, reusable software architecture for the querying and abstraction of temporal patterns using ontologies and rules;(2) to develop a data-mapping tool that can map time-oriented data stored in existing relational databases into a temporal ontology suitable for knowledge- based querying;and (3) to create a query-elicitation tool that can allow investigators to formulate domain- relevant temporal patterns for data extraction and inspect the results of those queries. We will develop and evaluate the use of these tools in ongoing collaborations with investigative teams who have well-developed data repositories in the areas of HIV drug resistance research and immune disorder trial management. We plan to undertake a reiterative process of software testing and optimization to ensure accuracy and performance. We will create a website to disseminate these tools (as open-source plug ins to Protigi). The widespread use of such general methods among investigators may greatly enable the discovery of scientifically relevant associations and patterns hidden among time-oriented data within biomedical databases.