• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Finally, you can manage your Google Docs, uploads, and email attachments (plus Dropbox and Slack files) in one convenient place. Claim a free account, and in less than 2 minutes, Dokkio (from the makers of PBworks) can automatically organize your content for you.


Introduction notes

Page history last edited by PBworks 13 years ago

 Definition and Orientation (Sarah)


 - There's nothing inherently wrong with conventional search engines that present the user with a list of results. But the more complex the search, the more difficult it becomes to manage the list and find the result you need. A simple list of web pages containing the search terms doesn't provide us with much insight.

 - The text list stems from the fact that most developers tend to have left-brain orientation (logical minded and a preference for text). However, the human mind grasps visual representations more quickly than an equivalent amount of text.

 - A number of search engines and tools have been developed over the last several years that use visual elements for searching and displaying results.

 - Much of the work being done on search engines is computer-centred - concerned with the inner workings of search: algorithms, optimizations, spidering and so forth.  Visual search engines are user-centred in that they support the needs of the user and how they think.

 - These can simplify the searching process, bring more meaning and context to the results, and allow users to view and conceptualize results in a different manner.


 - There are many different types of visual search engines, each using a visual element in a different way.  (This is not an exhaustive list. And some of these belong to more than one category.)


Clusters results based on context and meaning










Displays relationships between items in a group such as movies or artists




Uses pictures as “search terms”

*Like.com (shopping)


Screenshots of results



Managed Q


Visual search for images




Displays how sites are linked



 - Most of the new products, as well as most of the research has tended to focus on the clustering visual search engines. So that is what we’ll be focusing on. But we’ll also demonstrate examples of a few of the other types.


 - Clustering search engines provide a compact visual overview of the search results in the form of groups, maps, or other arrangements that display topics and themes in the results. Instead of scanning a list of results sequentially based on their importance as ranked by the search engine, searchers can see the larger view of the results and get a sense of how the different subtopics fit together, and which subtopics might be most useful.

 - Clicking on one of the topical clusters can greatly increase the precision of the search by gathering only results with the intended meaning or context.

 - For example, if a user types the word 'car', a traditional search engine would return a list with manufacturers, rental agents, mechanics, collectors etc. Users looking for rental cars would be very disappointed to find a list of manufacturers and collectors. However, with a visual search engine the user could click on the “rental agents” cluster to retrieve more precise and highly relevant results.

 - The difference between a traditional search engine and a visual search engine is like the difference between reading a book from front to back with the text organized by perceived importance, or looking at the table of contents or index first.


 - Most visual search engines are metasearch engines, summarizing the results of other search engines

 - Visualization tools use two basic approaches to clustering information: (1) they use metadata that is associated with the page. (2) they use statistical and/or linguistic algorithms to determine the meanings behind the words and the key concepts on the page. The results are then graphed visually according to the rules of the algorithm.



Key Concepts, individuals, and milestones related to visual search engines (Naz)



- Visual search engines are something more than fancy User Interface (UI). Different UI do not really change the results pages but only improve the user interaction with the query box.

- In digital environment, it is important that the user be able to take advantage of both keyword searching and classification structure. In the classic search engine, the meta-index had nothing to do with classification per se. In order to have a more efficient information retrieval system; we require a smart content organization system as well.


Conceptual framework and type of visulalization software realted to visual search engines:


1. Text clusters


The combination of Content Integrators and Clustering Engines are being used to automatically organize search or database query results into meaningful hierarchical folders, cluster them by subject categories using natural language.  (e.g. Clusty) 


2. Hyperbolic browser/ Hyperbolic Tree TM


It is a Focus+Context technique for displaying large tree structures. It provides browsable links that are context-relative and follow underlying semantic relationships. Now marketed by InXight, a Xerox offspring - the hyperbolic tree is now called Star Tree. (e.g. xrefer)

For pic visit: http://www.inxight.com/VizServerDemos/demo/orgchart.html



3. Information maps


VisualNet is software which transforms data from database into an information map that can be displayed through any Web browser. Using cartographic techniques to map search results from existing hierarchies and their subsidiary cluster (such as LC classification), VisualNet tend to be very stable and the clustering predictable.

Have two pics



4. Graphical


Using content management tools, such as Groxis, this is a graphic information delivery provider for some search engines. It is a multilevel application with server-side and client-side component.  (e.g. Grokker)


For pic visit: http://www.groxis.com/service/grokker/publishing.html



5. Hybrid visualization


It used by many libraries around the world, Medialab is a type of Hybrid visualization. The results are displayed in a typical browser search list by relevance to the query terms. In addition to a search list, the software has the ability to compares a user's search terms to the metadata in the library catalog to create a visual map (a "word cloud") of associations or overviews of areas of interest. These words are suggestions for patrons to discover new information and help them formulate a query. (e.g. AquaBrowser Library)

For pic visit: http://www.medialab.nl/index.asp?page=aquabrowserlibrary/overview


Comments (0)

You don't have permission to comment on this page.