Paper Reading #20: Designing a thesaurus-based comparison search interface for linked cultural heritage sources

Comments
Evin Schuchardt
Luke Roberts

Reference Information
Title: Designing a thesaurus-based comparison search interface for linked cultural heritage sources
Authors: Alia Amin, Michiel Hildebrand, Jacco van Ossenbruggen, Lynda Hardman
Presentation Venue: IUI 2010: Proceedings of the 15th international conference on Intelligent user interfaces; February 7-10, 2009; Hong Kong, China

Summary
In this paper, the authors discuss comparison search, a task in which people search for similarities or differences between sets of data. They focused on the cultural heritage domain. After doing a preliminary study on experts trying to compare sets of artworks, the researchers developed LISA a tool for cultural heritage experts doing complicated comparison searches of artwork.

Image of the LISA interface taken from the paper

Their paper mainly discussed the interface of the system and the interaction with it. They recommended reading another paper for information on the technological infrastructure.

In their design they focused on several selection and comparison challenges:
1. Searching artworks
2. Selecting artworks
3. Comparing artworks

For searching, the researchers included in their interface a place to type an artwork’s property (artist’s name, the date, the material used in the piece, etc.) and included an autocomplete feature to assist the user. They also assured that their system supported name aliases.

When selecting artwork, the user can include more than one piece of art in their set. They can drag and drop the thumbnail or select all search results at once and insert them into a set.

Currently their system only allows for two sets to be compared. Their system supports both single property (represented with a bar chart) and dual property (represented with a scatter plot) comparisons. Some of the properties compared include material, subject, height and date.

The researchers also mention that they implemented LISA on top of ClioPatria, a web-based application for searching through heterogeneous sets of data. They refer the reader to other papers for further information on this and information related to the dataset used.

They also did a user study to test the three features mentioned above. Different people affiliated with museums, art historical archives and universities participated in the study, which took place in their work places.

They compared the ease of use of LISA with RKDimages, a widely-used online cultural heritage archive. They found that overall LISA was easier to use than RKDimages, especially when it came to searching for many artworks and selecting artworks. They also found that the number of artworks searched for and selected did not degrade LISA’s performance. Overall, the participants preferred LISA over RKDimages. They appreciated the different ways to display the data. However, they wanted to know how the graphs were made and what they were based off of before they would trust the results.

Discussion
I found the paper interesting and the topic unique. I don’t know much about artwork or what researchers do when they search for artwork and do research with it, but the participants in the study did seem to appreciate the system. I also liked their interface. It looked easy to use and looked self-explanatory. While I am curious to know the more technical side of things, I did appreciate how in depth they went with their studies and explanations on why they designed the interface the way they did.

Perhaps I missed it, but I don’t think the authors ever mentioned what LISA stood for.

For future work they mention that they would like to improve the autocompletion feature, provide bookmarking, provide a search history feature and allow for more interaction with the artwork to enlarge it on demand.