Information Visualization for Searching and Browsing: Theories and Applications

Information Visualization for Searching and Browsing: Theories and Applications, ASIST Annual Meeting 2004

Javed Mostafa, Indiana University
Allison Druin, University of Maryland
Xia Lin, Drexel University
Jin Zhang and Dietmar Wolfram, University of Wisconsin-Milwaukee
Moderator: Jin Zhang

Javed Mostafa started out by showing screen shots of search engines using visualization. How do we get users to go deeper into the search? he wonders. How can we create a live interface to a live resource? He works on the Enable project at Indiana University. He mentioned scatter-gather, but he didn’t explain what it is. Mooter supposedly maps the data, then tries to learn based on what the user selects. It tries to remember the information to refine the search. A screenshot shows a map of books dealing with conversative and less conservative ideas. (I’m on the floor in the back near a power outlet (if only my batteries recharged as easily!), too far away to read any of the titles.)

Allison Druin shared some live demos with us, focusing on search tools for children 3 to 13 years of age. She works with the International Children’s Digital Library. Creating tools for children has different challenges than creating search tools for adults. Kids don’t visualize information like we do and they don’t like to type long URLs. Kids don’t always know where to look. Things that are obvious to us aren’t obvious to them. Children also have problems using search results in a standard fashion. The ways they like to approach and explore the results varies.

“Wow, cool! Look at how many books I got! So where do you think is the book I want?” said a child involved in one of her projects.

They gather research from a variety of places, including weblogs. They also work with international children. The library has thirty different languages.

The interface has lots of images. Allison describes the categories as “not your grandmmother’s library categories.” One of the neat search functions is the ability to find books by color. It’s very easy to find all of the red books. It’s also possible to find books in or about a country. She said kids have a difficult time using the engine to find books.

Multiple book readers are part of the library because kids want to do different things with the books. Some kids want to flip through the pages before selecting a book. Some want to start in the middle. Parents might want to read from the beginning to the end. It’s also possible to change the background colors in at least one of the readers (the Spiral).

Allison discovered 2/3 of the library’s users are adults, so they have to cater to both audiences. While images worked well for many of the kids, the adults wanted the text standard search engines have.

Kids thought the icons in the search results were confusing because they were square, like the icons for book pages. They found hierarchies and having to page through search results confusing.

“It’s not just searching, it’s how we use the stuff,” Allison expounds.

Q: What if a kid wants to find a similar book?
A: They’re their working on ways to group similarities, perhaps by user created categories. One way they’re thinking of apply this data is by adding information about the way books make someone feel, e.g. “This book makes me feel happy.”

Q: Do they work with international children?
A: Yes and they have grants to do it, too.

Q: I couldn’t hear the question. Allison paraphrased it by saying it’s about indexing.
A: Allison talked about international indexing. One of the challenges is that most of the research about children using search tools has been done in the United States. That research may not be appropriate for children overseas.

Q: What about collaborating with other children’s service providers, like Sesame Street?
A: We collaborate with a number of people. One of our researchers came from Sesame Street.

Xia Lin talked about visual mapping and visual mapping research. He’s interested in using the power of visualization for information seeking. He thinks it has great potential, but there’s skepticism still.

The Brain. Touchgraph. Grokker.

VisualLink and Visual Concept Explorer

VisualLink uses a global mapping feature. Localized mapping is for interactivity. It presents a user with a concept map of search results.

Visual Concept Explorer does similar things, but with a slightly difference appearance. When a user selects a term, the map changes to reflect the selection.

A questioner mentioned Highwire.

Dietmar Wolfram presented study results about how indexing characteristics change a document space.

One of the challenges of presenting at this conference is that there is no Internet connectivity for speakers. I knew about that before preparing my presentation. I could have done so much more had I had a net connection. Several times during the questions and answers, we could have provided better answers had there been a net connection because we could have shown someone something instead of just talking about it. One of the speakers on this panel didn’t realize she wouldn’t have a net connection and said so during her presentation. She changed her approach because of it.

Be Sociable, Share!