Week 13 May 3, 2011: Guest Lectures

Printer-friendly version

Guest Lecturer: Tim Clark http://madrc.mgh.harvard.edu/timothy-w-clark-ms
Title: Dynamic Semantic Metadata on Web Resources in Biomedicine [Download]
Guest Lecturer: Alyssa Goodman https://www.cfa.harvard.edu/~agoodman/
Title: The Visual Display of Information: The Art of Numbers"

Reading Assignment for this class:

Written Assignment for this class:

  • Formulate and post a question for each reading on this page. Include in your posting: Your name, your 2 questions and for each question, identify the paper for which it is a question.
  • Review the learning objectives you turned in and add a new section that describes the criteria by which you should be judged against these objectives

Presentation Assignment for this class: none

Name Questions
Eric Question for Alyssa: Many modern "modular" tools for data visualization exist for the web browser, and are depending on client-side computation of data provided in a specific format. One of the primary reasons for this is the availability of tools like Javascript and Flash that are compatible across operating systems and web browsers. This is fine for visualizations of small data sizes, but as your paper states, the trend in science is towards big data. How do you propose the design of tools that are modular at the server-side, close to the data, where you run into challenges like heterogenous software environments and data formats?
Question for Tim: What is your criteria for "innovation" computing?
Jin Question for Alyssa:One of the main problems discussed in the paper is data visualization. In the section "modular craftmanship", the paper mentioned creating interoperable and module-able data visualization tools for high dimensional, graphical displays of data. I think this interoperable and module-able data visualization tools are beneficial, but since software and tools may not be compatible or working well with each other, do you think there will be a generalized solution for interoperability problem? Since data can be in heterogeneous format also, how can we solve this problem in visualization?
Question for Tim: After reading the whitepaper, I am particularly interested in two sections: data provenance section and large-scale distributed scientific collaboration via internet. In both sections, you describes some benefits, use cases, and identify some issues in the area. So I am interesting in if you are considering using Semantic Web technologies to solve problems relate to provenance, distributed access to data, and integration of knowledge system? If so, how would you use semantic web technologies?
Ping Question for Alyssa: The introduction of your paper state that some people believe that computers can find patterns in quantitative information better than people can, some people do not believe so. Which of the two groups are you in? If you do not believe computers are better at finding patterns, could you tell me your reasons and share your experience of dealing with data as a scientist in the discipline of Astronomy?
Question for Tim:I have some questions about the SWAN project. How do you evaluate the impact of SWAN? Do you collect feedback from the users? How do you get scientists and researchers use SWAN?
Scott Question for Alyssa:Are there visualization types or techniques that can work in virtually any discipline (beside the ones we already use)? If not, do you see us going back to each visualization being a custom work?

Question for Tim:Is the IIC on track for the "2011 and beyond" portion of the Structure and Evolution? If so, have things gone mostly to plan? If not, what changes were necessary?
Peter Question for Alyssa:No Questions. Very interesting work.