CPE/CSC 481-W11 Knowledge-Based Systems Project Ideas


Visualization of Ontologies and RDF Graphs

The underlying data structures for ontologies and RDF (Resource Data Framework) structures are graphs. While graphs in principle are well suited for visualization purposes, there are several challenges for these particular structures. Ontologies frequently have a multitude of relationships (links in the graph), and it is impractical to display all of them simultaneously. RDF graphs frequently are huge, and can’t be displayed in their entirety. In both cases, however, there are underlying principles that humans intuitively utilize to arrange the nodes in both a visually pleasing and easy to comprehend manner (although it may be very difficult or impossible to achieve both). The objective of this project is to examine existing visualization methods for ontologies and RDF graph, and identify their advantages and shortcomings. On the other side, there are packages available for displaying generic graphs. Our hope is that a combination of insights from the existing visualization tools with advanced graph display packages, together with heuristics that reflect organizational principles for ontologies and RDF graphs will result in better visualization tools. Ideally, a prototype of such an ontology visualization tool would be integrated with the Protégé ontology development environment.

Computers and Knowledge Educational Resources

The goal of this project is to provide educational resources for a course similar to this one, but intended for participants without a strong computer background.

iFixit Recommender System

The ifixit Web site offers instructions for repairs of all kinds of devices, and has recently been expanded to allow the submission of user-generated instructions. The company, co-funded by a CSC alumnus, is interested in adding a recommendation facility that would offer user suggestions such as other devices or repairs they may be interested in, or parts and tools that are suitable for the repair. They currently have recommendations for the latter, but they are not very sophisticated.

Knowledge management system for topic analysis in office environments

The goal of this project will be to produce a knowledge management system to support the mining of data in typical office communication environments to answer questions about what people are working on and what deliverables they are producing. For example, who worked on robot coordination algorithms in a given month?
- social networking
- latent semantical indexing
- query communications analysis data
- Python script to generate test dat from Wikpedia
- plugin architecture for visualization
- framework of entities and relations
- fits into several projects
- emphasis could be on visualization or analysis

Contact persons

- Jeanine Ayers, Nathan Schurr; Aptima

Background required


- statistics, model checking, machine learning
- at least one of {Flex, REST/SOAP, Web-based development}

Decision Maker Modeling

The goal of this project will be to more accurately model decision makers and the allocation of their tasks across the domain of air traffic control. The decision makers have a profile (current task load, task of focus, capability matrix) and the system must make decisions to meet merging and changing goals within the environment such as optimizing humans, systems, safety, etc.
- general resource allocation problem
- modeling of decision makers, e.g. situational awareness, task switching
- data gathering to populate the model (more flexible, but possibly more challenging)
- ontology (possibly integration with existing agent code)

Contact persons

- Nathan Schurr; Aptima

Background required


- ontology, JADE

Knowledge management system for operator's understanding of robot movements

The goal of this project will be to produce a knowledge management system to support the operator's understanding of the movement of a team of robots through a building layout. This system will account for the ability of (1) the robot team to make decisions about the level of control it exhibits; and (2) decisions the team of robots make about when and what needs to be communicated to the human operator of the team.
- multi-robot coordination
- phase 2 of a research project
- visualization of the status of the robots
- tradeoff between autonomy and
- very open
- open source environment ROS, Willow garage; http://www.willowgarage.com/blog/2010/07/22/urbi-open-source-now-integrated-ros
- possibly involve 484 students, Tracy Davies
- Xbox Kinect
- also Brown University, iRobot

Contact persons

- Jeanine Ayers, Nathan Schurr; Aptima

Background required

- basic robotics

Project Title

Project Description