• Search |
• Edit this page

Old Capstone Paper

For Combining Visualisation and Interaction to Facilitate Scientific Exploration and Discovery Workshop:

Questions:

• How can interactive visualisation methods and tools be augmented to address concerns of both scientific computing and HCI?
• Which research questions need to be considered when aiming to achieve efficient HCI and 'human-to-human' interaction between scientists while exploring complex data spaces?
• How can we define effective abstractions for the visualisation and user interaction processes?
• What is the impact of the task or application field on interactive visualisation?
• To what extent are usability problems independent of the context of use and need to be taken into account when designing interactive visualisation environments?

Topics:

• Advanced visualisation algorithms and systems;
• Adaptive interaction for effective data exploration and knowledge discovery;
• Human factors in computer-supported exploration;
• Interactive visualisation applications and systems.

Paper:

Opening a Window in the Black Box: Improving bioinformatics tools by exposing their innards to biologists

Biologists are using more and more computational tools in their work, and understanding these tools is critically important to doing good science. However, because biologist's research focus is biology and not computer science, they are motivated only to learn "just enough" about a computational tool to perform the operations they immediately need. As a result, biologists sometimes use tools incorrectly, and fail to see analyses that would be potentially helpful to their research.

Many resources are available to help biologists, including context-specific documentation, books, papers, interactive demonstration applets and courses on bioinformatics software. Unfortunately, most of the biologists I talked to made minimal use of these. In general, biologists were willing to only ration a small amount of effort towards understanding computational tools, as they see these tools as a small part of their work. Tragically, when they do invest this time, they are often rewarded with unhelpful documentation written in confusing jargon.

Underbelly is a sequence alignment tool that has been designed to expose its internals to biologists during its use. Typically, bioinformatics software provides minimal feedback to users while it is running. Underbelly shows the users the data being manipulated at high speed, allows them to pause the action, and move back and forth through the running history of the algorithm.