Visual search for catalogs, collections and archives

Abstract

Searching is a necessity. Search should be fun. Catalogs, collections and archives contain large amounts of data ready to search. But the most common presentation of the results in the form of long lists of text seems to be rather repulsive. Various studies have shown that only very few of these results attract the users’ attention. Improving the order of search results is therefore at the center of numerous research projects. Visual search is an alternative approach based on an observation from the pre-digital era that avoids these technical complexities: extensive image libraries – the slides on a light table – allow for easy scanning and ordering of large result sets. Visual search aims to bring this approach to the web. The authors first followed this concept while implementing oSkope, a visual search engine through web services such as Amazon, eBay or YouTube. The key design patterns were suitable for reuse in a number of Rich Internet Applications for collections and product catalogs since realized. By combining an intuitive and fluid navigation with a clear, meaningful and aesthetically pleasing presentation, these applications attract the users’ attention to a larger set of results. Additional information can be displayed on demand. Sophisticated search options allow for an accurate search. The results can be saved for later use or be sent to a friend.