Large-Scale Querying of Semantic (RDF) Data
The proliferation of the semantic web in the form of Resource Description Framework (RDF) demands an efficient, scalable, and distributed storage along with a highly available and fault-tolerant parallel processing strategy. More precisely, the rapid growth of RDF data raises the need for an efficient partitioning strategy over distributed data management systems to improve SPARQL query performance regardless of its pattern shape with minimized pre-processing time. In this context, we are investigating new relational partitioning schemes for RDF data, that further partitions existing Property Table into multiple tables based on distinct properties (comprising of all subjects with non-null values for those distinct properties) in order to minimize input data and join operations. Evaluation of these techniques involves extensive experimental evaluation with respect to preprocessing costs and query performance, using various benchmark datasets such as Lehigh University Benchmark (LUBM) and Waterloo SPARQL Diversity Test Suite (WatDiv) datasets with varying number of triples (i.e, large-scale datasets in the order of billions of triples).
Information or data sharing among different systems is very limited due to the heterogeneous nature of data in syntaxes, structures and semantics. An ontology is a formal description of knowledge as a set of concepts and relationship between them which represents a specific domain. Ontology matching is a process of finding semantic relationship between two or more different ontological entities. Ontology matching is an integral part of creating Linked Data which is a process of publishing and linking structured data on the web. It is used to overcome the limitation of semantic interoperability between current vast distributed systems available on the internet. Despite the advancement in Linked Data, the ontology matching is still mostly done manually by domain experts which is labor intensive and error prone. Ontology matching or linking between source and target ontology is generally done on two different levels. The first one is schema level and the second one is instance level. Schema level matching is mapping/alignment between concepts/classes of the source and target ontology and on the other hand, instance level linking is mapping/alignment between instances/individuals of the source and target ontology. There is another type of matching which is mixed of schema level and instance level of linking. We investigate schema level linking of ontologies and evaluate our approach using datasets published in Ontology Alignment Evaluation Initiative (OAEI).
Semantic ETL framework for Big Data Integration
Big Data researchers are dealing with the Variety of data that includes various formats such as structured, numeric, unstructured text data, email, video, and audio. The proposed Semantic Extract-Transform-Load (ETL) framework that uses semantic technologies to integrate and publish data from multiple sources as open linked data provides an extensible solution for effective data integration, facilitating the creation of smart urban apps for smarter living.
Question Answering over Linked Data
Most question answering (QA) systems over Linked Data, i.e. Knowledge Graphs, approach the question answering task as a conversion from a natural language question to its corresponding SPARQL query. A common approach is to use query templates to generate SPARQL queries with slots that need
to be filled. Using templates instead of running an extensive NLP pipeline or end-to-end model shifts the QA problem into a classification task, where the system needs to match the input question to the appropriate template. We investigate approaches to automatically learn and classify natural language
questions into corresponding templates using recursive neural networks. We evaluate our system using benchmark datasets such as LC-QuAD and QALD challenge datasets.
CircuitTutor (Co-PI)
Circuit Tutor is a step-based computer-aided tutoring system to aid in the teaching of elementary linear circuit analysis. This project involves a unique approach to computer-aided instruction where the computer both creates and solves its own circuit problems on the fly using a sophisticated three-step algorithm. Each student can access an unlimited source of fully worked and explained examples at various levels of gradually increasing difficulty, as well as exercises designed to be completely isomorphic to those examples, but completely different both in circuit topology and element values. No two students work the same problems, eliminating situations where one student merely copies the work of another or that in a solution manual, since none exists. Both problems and solutions are error-free, eliminating the frequent frustration students experience when working conventional textbook problems and examples, where errors are notoriously common. Moreover, exercises are presented in a game-like format to significantly improve student motivation and engagement; students even frequently report that they are ‘fun’! A variety of special pedagogical features such as color-coding of nodes is used to enhance student learning.
External Funding: Improving Undergraduate STEM Education: Education and Human Resources (IUSE: EHR) Program of the National Science Foundation under Grant No. 1821628 August 2018 – July 2022.
Instructional Module Development System (IMODS)
Description: Design and development of a framework for outcome-based course design process and its translation into a Semantic Web-based software tool that can guide STEM educators through the complex task of curriculum development and provide relevant information about research-based pedagogical and assessment principles.
Area of Study: Semantic Computing, Engineering Education, Ontology Engineering, User-Centered Design
External Funding: National Science Foundation’s Transforming Undergraduate Education in Science, Technology, Engineering and Mathematics (TUES) program Award No. DUE-1246139; July 2013 – June 2016.
Recent Publications
- S. K. Bansal, J. M.-Alexander. “Building a Repository of Instructional and Assessment Techniques for Instructional Module Development System”, in Proceedings of Research in Engineering Education Symposium (REES), pp. 771-780, Cape Town, South Africa, 2019.
- V. Raj*, P. Goulet*, S. K. Bansal. “Evaluation of Instructional Module Development System”, in proceedings of IEEE Frontiers of Education Conference (FIE), pp. 1-9, San Jose CA, 2018.
- S. K. Bansal, O. Dalrymple. “Repository of Instructional and Assessment Techniques for OBE-based Instructional Module Development System“. In Journal of Engineering Education Transformations (JEET) Volume 29, Issue 3, 2016, pp. 93-100.
- S. K. Bansal, A. Bansal, O. Dalrymple. “Outcome-based Education Model for Computer Science Education“. In Journal of Engineering Education Transformations (JEET) Volume 28, Issue 2&3, 2015, pp. 113-121.
- O. Dalrymple, S. Bansal, A. Gaffar, R. Taylor*. “Instructional Module Development (IMOD) System: A User Study on Curriculum Design Process“. In Proceedings of IEEE Frontiers in Education Conference (FIE), October2014, Madrid, Spain.
- O. Dalrymple, S. Bansal, A. Gaffar. “User Research for the Instructional Module Development (IMOD) System“. In Proceedings of American Society for Engineering Education Conference (ASEE) – NSF Grantees session, June 2014, Indianapolis, USA.
- O. Dalrymple, S. Bansal, K. Elamparithi*, Husna Gafoor*, Adam Lay*, Saishruti Shetty*. “Instructional Module Development (IMOD) System: Building Faculty Expertise in Outcome-based Course design“. In Proceedings of IEEE Frontiers in Education Conference (FIE), October 2013, Oklahoma City, USA.
