Abstract 1 1 Introduction 2

Download 324.2 Kb.

Page	15/20
Date	10.08.2017
Size	324.2 Kb.
	#31133

1 ... 12 13 14 15 16 17 18 19 20

Data mountain
ET MAP
FilmFinder
Galaxies of News, Personalized Galaxies of Information

Connex

Reference: Beth Hetzler, W. Michelle Harris, Susan Havre and Paul Whitney. Visualizing the Full Spectrum of Document Relationships. Structures and Relations in Knowledge Organization. Proc. 5th Int. ISKO Conf. ERGON Verlag, 1998, pp. 168-175.

URL: http://multimedia.pnl.gov:2080/infoviz/technologies.html
Connex is a prototype visualization tool that displays relationships among a collection of entities. The X and Y axes correspond to an ordered list of entities. The Z axis corresponds to various relationships among entities. Color is assigned to categorize relationships. For multiple relationships, layering may be applied.

Data mountain

Reference: Robertson, G., Czerwinski, M., Larson, K., Robbins, D., Thiel, D. & van Dantzich, M., Data Mountain: Using Spatial Memory for Document Management, In Proceedings of UIST '98, 11th Annual Symposium on User Interface Software and Technology, pp. 153-162.

URL: http://research.microsoft.com/users/marycz/uist981.doc

Data Mountain is used as a “favorites” mechanism of Web browser. It presents a collection of favorite Web pages by miniature pictures of a Web page itself as an icon image. The placing of icons is arbitrary in 3D by user. The usability study of Data Mountain showed that storing information using spatial location improved performance in retrieving information. Subjects indicated that they knew the location of specific favorite pages.

ET MAP

Reference: Chen, H., Schuffels, C. and Orwig, R. Internet categorization and search: a self-organizing approach. Journal of Visual Communication and Image Representation 7(1), 1996, p. 88-102.

Chen, H., Houston, A.L., Sewell, R.R. and Schatz, B.R. Internet browsing and searching: user evaluation of category map and concept space techniques. Journal of the American Society for Information Science 49(7), May 1998, p. 582-603.

URL: http://ai.bpa.arizona.edu/start.html

ET-MAP is an application based on a multi-layer Kohonen self-organizing feature map. ET MAP has been used to catalog Web pages. The input documents are used for training a Kohonen Self Organizing Map (SOM), similar to SiteMap. After an initial map is developed, the process is recursively applied to the areas which contain high numbers of documents as layer map.
The ET-MAP implementation provides a whole map view. Zooming into selected areas shows deeper layers of the map view or documents list view. The document list view is a hypertext link to the original Web page.

FilmFinder

Reference: Christopher Ahlberg and Ben Shneiderman. Visual Information Seeking: Tight coupling of dynamic query filters with starfield displays. Proc. of ACM CHI94 Conference (April 1994), p. 313-317 + color plates.

URL: http://www.cs.umd.edu/hcil/spotfire/
FilmFinder is a prototype, developed by HCIL lab, that demonstrates the concept of dynamic queries. Other prototypes include HomeFinder and Spotfire. The dynamic queries approach applies the principle of direct manipulation to the information seeking task in a database environment. The key principles of dynamic queries are;

visual presentation of the query’s components,

visual presentation of results; rapid, incremental, and reversible control of query,

selection by pointing, not typing, and

immediate and continuous feedback.
In FilmFinder, films are presented as small, colored rectangles on an XY plot. X axis represents time and Y axis represents a measure of popularity. Color encodes film categories. Dynamic queries are applied by having control components, i.e. sliders, check boxes, and buttons, to control which films are visible.
The display area allows zooming. When the number of films displayed is small, labels are automatically added. Clicking on the film causes a “detail-on-demand” window to pop up.

Galaxies of News, Personalized Galaxies of Information

Reference: Rennison, E. Galaxy of News: An Approach to Visualizing and Understanding Expansive News Landscapes. UIST 94 ACM.

Rennison, E., Personalized Galaxies of Information. Proceedings of the Conference on Human Factors in Computing Systems 95, Denver, Colorado. May, 1995.

URL: http://www.acm.org/sigchi/chi95/proceedings/demos/er_bdy.htm

Galaxies of Information is a system that provides mechanisms to analyze the relationships between information objects. It builds a representation of the underlying structure of the entire body of information. The current implementation shows the relationships between information using an Associative Relation Network, ARN. The documents are used to establish the weights of nodes of the ARN, words extracted from the documents. The information hierarchy is extracted from the ARN by statistical dependency rules.
The interface, similar to PAD, uses 2D space with unlimited depth. A hierarchy of keywords is placed at different levels. As a user zooms in to get more detail, the upper level fades away. Unlike PAD, documents may be in many places, since a single document can be pointed to by many keywords.
The system is currently applied to news. Given the nature of news, information changing over time, the structure in this scheme must change dynamically. Personal Galaxy of Information uses a learning algorithm to adaptively customize the presentation of information to a particular user's interests. The user can specify a preference or bias for news information structure. User preferences are learned when a user navigates through the information space and reads a document. The preference information may be reinforced, i.e. indicating dislike for a document or adjusting learning parameters.

Download 324.2 Kb.

Share with your friends:

1 ... 12 13 14 15 16 17 18 19 20