Graphical Interfaces to Support Information Search

Main | Overviews | Classics | Demo Potential | Other systems | User-testing | Bibliographies

Other Graphical Interface Systems of Interest


Two Dimensional Systems | Three Dimensional Systems


Envision

Envision

Nowell, L. T. (1996) Visualizing Search Results: Some Alternatives to Query-Document Similarity. In Frei, H.P. et al., (Eds.) Proceedings of the 19th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '96) (pp 65-75). New York : ACM.


.pdf available from ACM Digital Library-SIGIR'96 or
http://opac3.cc.vt.edu/SIGIR96.Env.html (recommended)

Lucy Terry Nowell, Robert K. France, Deborah Hix, Lenwood S. Heath and Edward A. Fox, Virginia Tech Department of Computer Science and Computing Center

Author's Abstract
A digital library of computer science literature, Envision provides powerful information visualization by displaying search results as a matrix of icons, with layout semantics under user control. Envision's Graphic View interacts with an Item Summary Window giving users access to bibliographic information, and Xmosaic provides access to complete bibliographic information, abstracts, and full content. While many visualization interfaces for information retrieval systems depict ranked query-document similarity, Envision graphically presents a variety of document characteristics and supports an extensive range of user tasks. Formative usability evaluation results show great user satisfaction with Envision's style of presentation and the document characteristics visualized.

Additional Comments
Envision gives users control over the layout of the graphical presentation. The users choose which variable to plot on the x-axis and the y-axis, as well as whether the icon size and shape should represent additional variables. Despite the wide number of choices, the display is still simple and not too flashy. Envision has problems representing several documents that should appear in the same place - it shows a single icon with the number of overlapping documents in the center. This means that the other features of those documents are obscured in the graphical representation. The developers of the system interviewed 12 professionals in the HCI, LIS and CS fields to determine what features they would find useful in a graphical interface - this groundwork seemed to pay off, since users responded well to Envision when it was actually tested. Does the system provided some sample/default displays, especially for first time users? This would make the system much more usable for the novice. While repeat/frequent users can certainly benefit from the control and choices provided by Envision, default display options could also make the system useful for novice users. The article makes references to many other visualization systems.


VQuery

VQuery

Jones S. (1998). Dynamic query result previews for a digital library. In I.H. Witten, R. Akscyn & F.M. Shipman, III (Eds.), Proceedings of ACM Digital Libraries '98,(pp 291-292).


.pdf available from ACM-DL'98
http://www.cs.waikato.ac.nz/~nzdl/publications/1998/Jones-Dynamic-DL98.pdf

Author's Abstract
Previous models of dynamic querying supported by query previews have focussed on attribute based querying, have required information providers create preview tables, and have provided little information to support initial query refinement. We present an alternative model that has been implemented for the New Zealand Digital Library, and describe the system architecture and user interface.
Users of information retrieval systems are conventionally locked into a querying cycle of: form query - submit query - wait for results - assess results - refine query - submit query and so on. There are two bottlenecks in this cycle. The first occurs when the user waits for a response from the information server. The second occurs when users consider the results and decide how to refine the query. Static, segmented presentation of result sets does little to support assessment and query refinement. Two techniques have been proposed to remedy this situation. Query previews store metadata (such as titles, dates, authors and so on) extracted from the information space in rapid access data structures to provide immediate response to user queries. Dynamic queries provide users with immediate updates of displayed result sets whenever query attributes are changed.

Additional Comments
This brief article provides an introduction to dynamic queries, explaining their potential benefits to the user. The VQuery interface could benefit from user-testing, but the underlying ideas are sound. Graphical dynamic queries are certainly an innovative interface for searching.


BibRelEx

BibRelEx

Bruggemann-Klein A, Klein R, Landgraf B. BibRelEx: exploring databases by visualization of annotated content-based relations. [Journal Paper] D-Lib Magazine, vol.5, no.11, Nov. 1999, Publisher: Corporation for National Research Initiatives,USA.

http://mirrored.ukoln.ac.uk/lis-journals/dlib/dlib/dlib/november99/landgraf/11landgraf.html
Also available at: http://wwwpi6.fernuni-hagen.de/Forschung/BibRelEx/pictures.html

Author's Abstract
Traditional searching and browsing functions for bibliographic databases no longer enable users to deal efficiently with the rapidly growing number of scientific publications. The main goal of our project BibRelEx is to develop a new method based on the visualization of content-based relations between documents such as cites, succeeds, improves with respect to. BibRelEx will therefore use these relationships for effective exploration. In addition, BibRelEx will take advantage of the additional insights into the area that can result from the aggregation of expert knowledge, which complements the specialized knowledge represented in the documents themselves. We are preparing to test this approach using a bibliographic database in a specific area of computer science.

Additional Comments
This two dimensional system is similar to several others, but the application to citation searching and relationships between citations is novel.


VOIR

Golovchinsky, G. & Chignell, M.H. (1997). The Newspaper as an Information Exploration Metaphor. Information Processing and Management, 33(5), 663-683.


Author's Abstract
The newspaper represents a mature information presentation medium that is well-suited to the display of relatively short, loosely related pieces of text. This work examines the implementation of the newspaper metaphor in an appropriate interface paradigm for large-scale full-text databases. Similarities between newspapers and hypertext databases lead us to suggest that this metaphor is appropriate for large automatically generated hypertexts, independent of the nature of their content. We describe VOIR, a software prototype that we have used as an electronic newspaper workbench. The program constructs newspaper pages interactively, and allows users to specify their information-seeking intent in a variety of ways, including graphical Boolean queries, hypertext links, and typed-in queries. We report some empirical evidence that indicates that users perform better when they are shown more articles simultaneously. These results suggest that the newspaper metaphor may be an effective organizing principle for a class of information exploration tasks. Finally, we discuss some implications that this work has for hypertext and information retrieval in general.

Additional Comments
The design of the VOIR interface is justified through the interesting similarities between full text databases and newspapers. Interestingly, the results of the study indicated that simultaneous displays of articles may help users perform better.


Snap-Together Visualization

Snap-Together Visualization

Snap-Together Visualization

Chris North, Ben Shneiderman, Anna Fredrikson, Gunjan Dang, Manav Kher, Human-Computer Interaction Lab, University of Maryland.

http://www.cs.umd.edu/hcil/snap/

Author's Abstract
Information visualizations with multiple coordinated views enable users to rapidly explore complex data and discover relationships. However, it is usually difficult for users to find or create the coordinated visualizations they need. Snap-Together Visualization allows users to coordinate visualizations to create multiple-view interfaces that are customized to their needs. Users query their relational database and load results into desired visualizations. Then they specify coordinations between visualizations for selecting, navigating, or re-querying. Developers can make independent visualization tools 'snap-able' by including a few simple hooks.

Additional Comments
Unlike many graphical information visualization projects, the Snap-Together Visualizations have been widely applied to many different types of applications. The designers' commitment to allowing the user to customize the system to suit their needs is one solution to the problem of creating useful and usable graphical information visualization systems. The project website includes several technical reports.


Scrollbar-based Visualization

Scrollbar-based Visualization

Byrd, D. (1999) A Scrollbar-based Visualization for Document Navigation. In Fox, E.A. et al., (Eds.)Proceedings of the fourth ACM conference on Digital libraries (DL '99)(pp 122-129). New York: ACM.


.pdf available from ACM Digital Library - DL'99

Author's Abstract
We are interested in questions of improving user control in best-match text-retrieval systems, specifically questions as to whether simple visualizations that nonetheless go beyond the minimal ones generally available can significantly help users. Recently, we have been investigating ways to help users decide - given a set of documents retrieved by a query - which documents and passages are worth a closer examination.

We built a document viewer incorporating a visualization centered around a novel content-displaying scrollbar and color term highlighting, and studied whether the visualization is helpful to non-expert searchers. Participants' reaction to the visualization was very positive, while the objective results were inconclusive.


Self-Organizing Maps

Self-Organizing Maps

Yang, C.C., Chen, H. & Hong, K.K.(1999) Visualization Tools for Self-Organizing Maps. In Fox, E.A. et al., (Eds.) Proceedings of the fourth ACM conference on Digital libraries (DL '99)(pp 258-259). New York: ACM.


.pdf available from ACM Digital Library - DL'99

Author's Abstract
Various statistical and pattern recognition techniques, such as concept spaces and category maps in the Illinois Digital Library project, has been explored to solve the semantic interoperability problem in DLI-1. Self-organizing category map is identified as a powerful tool for information summarization. However, visualizing large-scale self-organizing map in a restricted size of window is difficult. For smaller regions, displaying labels is infeasible. In this paper, two visualization tools, fisheye view and fractal view, are presented. It assists users to visualize a large-scale self-organizing map geographically and semantically.


libViewer: Visualizing the Library

libViewer: Visualizing the Library

Rauber, A. & Merkl, D. (1999) SOMLib: A Digital Library System Based on Neural Networks. In Fox, E.A. et al., (Eds.)Proceedings of the fourth ACM conference on Digital libraries (DL '99)(pp 240-241). New York: ACM.


.pdf available from ACM Digital Library - DL'99

Author's Abstract
Digital Libraries have gained tremendous interest with numerous research projects addressing the wealth of challenges in this field. While computational intelligence systems are being used for specific tasks in this arena, the majority of projects relies on conventional techniques for the basic structure of the library itself. With the SOMlib project we create a digital library system that uses a neural network-based core for library representation and query processing. The self-organizing map, a popular unsupervised neural network model, is used to automatically structure a document collection. Based on this core, additional modules integrate distributed libraries and create an intuitive representation of the library, automatically labeling the various topical sections in the document collection.


Nemo

Nemo

Hascoet, M. & Soinard, X. (1998). Using Maps as a User Interface to a Digital Library. In W.B. Croft, et al. (Eds.), Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. (pp 339-340).


.pdf available from ACM Digital Library - SIGIR'98

Author's Abstract
The Nemo project aims to develop a graphical interface contributing to fast navigation and exploration in the Digital Library of EDF. The digital library of EDF contains collections of documents in SGML-HyTime format, containing information on people, activities, projects, scientific work, etc. These documents are currently accessed through traditional Boolean query interface and the aim of the Nemo project is to provide customizable graphical maps of documents resulting from a query. The originality of the maps lies in the fact that (1) they are computed dynamically from the characteristics of the data retrieved with a query, (2) the mapping of the graphical attributes with data characteristics is highly and easily customizable and (3) map layout account for a perspective (or classification system) that is also customizable.

Additional Comments
While the system is not fully explained in this brief article, the ideas and the IR interface have potential. Each icon has 6 graphical attributes: shape of icon, background pattern, label, size, background color and foreground color, which can be configured to signify document properties like type, date, importance, etc.). A "specific graphical interface" has been developed to enable a user to specify graphically things like intervals of dates, etc. This system could be useful for expert searchers, since it is customizable, and the interfaces to change the assigned attributes seem understandable. Since the system requires user input to customize the graphical attributes, it would be difficult for novice searchers, unless a standard default or "novice interface" were developed to help them choose appropriate values for the graphical attributes.


Three Dimensional Systems | Two Dimensional Systems


NIRVE

NIRVE

Interactive 3D Visualization for Document Retrieval


http://zing.ncsl.nist.gov/~cugini/uicd/viz.html
John Cugini, Dr. Christine Piatko, Dr. Sharon Laskowski
Visualization and Virtual Reality Group
Information Access and User Interfaces Division
Information Technology Laboratory
National Institute of Standards and Technology (NIST)

Author's Abstract
The availability of large collections of documents coupled with powerful search and retrieval algorithms provides the opportunity for people to access large sets of relevant documents in electronic form. However, often a user query can result in hundreds of potentially useful documents returned by the system. Our hypothesis is that interactive 3D graphics techniques can be used to help the user comprehend and filter such result sets. We describe some prototypes developed at NIST in pursuit of these goals and discuss associated design issues, such as icon appearance, layout within 3D space, and interaction mechanisms.

Additional Comments
An early report on the development of the NIRVE system, outlining initial work on three design paradigms for visualization: a document spiral emphasizing relevance, a three-keyword axes display emphasizing frequency of each keyword, and a nearest neighbor cluster emphasizing relationships between documents. The novel idea of presenting multiple visualization options is described, and future work outlined.


NIRVE

Document Clustering in Concept Space: The NIST Information Retrieval Visualization Engine (NIRVE)


http://zing.ncsl.nist.gov/~cugini/uicd/cc-paper.html
John Cugini, Sharon Laskowski, National Institute of Standards and Technology (NIST)
Christine Piatko, Johns Hopkins University / Applied Physics Laboratory (JHU/APL)

Author's Abstract
The NIST Information Retrieval Visualization Engine (NIRVE) is a 3-dimensional interface designed to enhance the information retrieval process by providing an overview of a set of text documents as well as access to the details of the individual documents through seamless user navigation and manipulation. Users can redefine the display space by consolidating related keywords from the input query into concepts, and by indicating the relative importance of those concepts. Document clustering is based on these user inputs. It is our hypothesis that this combination of 3D graphics and fine control by the user provides a more effective interface than is found in many comparable systems.

Additional Comments
The NIRVE system relies on human user expertise to cluster the results, after automatic searching and clustering has been performed. The NIRVE system focuses on giving the user a powerful set of tools to manipulate their search results, including a NIRVE Control panel, with buttons arranged to guide the user through the typical steps of search refinement, a Concept Control window to display the combination and weight of current keywords, and a Document Space with a 3-D visualization of clustered search results. Each component is pictured and described in the report.


NIRVE

Design of 3-D Visualization of Search Results: Evolution and Evaluation


http://www.itl.nist.gov/iaui/vvrg/cugini/uicd/nirve-paper.ps.gz

John Cugini, Sharon Laskowski, National Institute of Standards and Technology (NIST)
Marc Sebrechts, The Catholic University of America

Author's Abstract
We discuss the evolution of the NIST Information Retrieval Visualization Engine (NIRVE). This prototype employs modern interactive visualization techniques to provide easier access to a set of documents resulting from a query to a search engine. The motivation and evaluation of several design features, such as keyword to concept mapping, explicit clustering, the use of 3-D vs. 2-D, and the relationship of visualization to logical structure are described. In particular, the results of an extensive usability experiment show how visualization may lead to either increased or decreased cognitive load.

Additional Comments
Many variations on the 3-D NIRVE display were created, and many sample screenshots are provided in the report. The NIRVE research is unique in the examination and testing of many different possible displays, all based on the same underlying system. NIRVE maintains a balanced focus on the relationship of the visualization to the other design features, rather than concentrating only a specific design or feature. This broader consideration of the benefits of visualization, and the continued research on the NIRVE project, make it both an interesting and influential system.


BookMap

BookMap

Navigation and interaction within graphical bookmarks
Mountaz Hascoet, University of Paris-sud, France


http://www-ihm.lri.fr/~mountaz/Publi/ri1232.pdf

Author's Abstract
This paper presents a new user interface called BookMap for WWW navigation tasks. This interface is an alternative to current browsing interfaces. The user's bookmarks are represented in a graphical document called a BookMap. The BookMap is enriches with othe web pages: (1) a short term history (the six last visited pages), (2) a personal "BestOf", (the six most visited web pages) and (3) an are of unclassified web pages stored to defer reading. A BookMap is considered both a document and a user interface for browsing.
In this paper, we first motivate our approach by reviewing previous analysis of web navigation problems and bookmark usage. We then quickly recall the concept of document as user interfaces. Then we describe the BookMap user interface and discuss visualization interaction and implementation issues.

Additional Comments
The information visualization techniques in this article apply specifically to the organization of personal bookmarks, but the added features greatly enrich the graphical interface.


BookMap

Intergration of navigational aids in the user interface.
Mountaz Hascoet, University of Paris-sud, France


http://www-ihm.lri.fr/~mountaz/Publi/ht00.pdfA user interface combining navigation aids, Hypertext'00, San Antonio, June 2000.

Author's Abstract
This paper describes how four independent navigational aids can be smoothly integrated in a unified graphical user interface. Our aim is to improve user performance in most web related tasks through integration of navigational aids. The navigational aids considered are (1) a short term history (the six last visited pages), (2) a personal "BestOf", (the six most visited web pages), (3) an area of unclassified web pages stored to be read later and (4) an overview of an organized collection of bookmarks.

Additional Comments
Some ideas from this system could be applied to information retrieval, especially in the case of full text articles or web-pages. The BookMap basically combines four existing techniques for managing visited webpages into a single system. The graphical interface, and the display of the four separate techniques simultaneously, are both quite interesting.


Shape-Based 3-D glyphs

Shape-Based 3-D glyphs

Rohrer, R.M., Sibert, J. & Ebert, D.S. (1999). A Shape-based Visual Interface for Text Retrieval. IEEE Computer Graphics and Applications, 19(5), 40-6.


.pdf available from IEEE Digital Library

Author's Abstract
"In this article we describe a shape-based visual interface for information retrieval and interactive exploration that exploits shape recognition. Our exploratory system uses procedurally generated shapes coupled with an underlying text-retrieval engine. A visual interface based on 3D shapes (glyphs) enhances traditional text-based queries and summarization. Our interface lets users visualize multidimensional relationships among documents and perceive more information than with conventional text-based interfaces. It promotes information overview and "drill-down" in support of analysis."

Additional Comments
Since this interface is still in the experimental stage, it has few practical applications, and searchers would need shape-recognition training before viewing their results. The system uses different shaped 3-D blobs to graphically represent varying degrees of relevance for particular search terms. While the 3D glyphs are an interesting idea, on paper the three dimensions are hard to imagine in this context, and the authors have difficulties justifying their use of 3-D instead of 2D in their explanation. The project could use user-testing to determine its usefulness and usability.


Islands

Islands

Brooks, M. and Campbell, J. Interactive Graphical Queries for Bibliographic Search. Journal of the American Society for Information Science, vol.50, no.9, Sept. 1999, 814-825.

Author's Abstract
This article presents "Islands," an interactive graphical interface for construction, modification, and management of queries during a search session on a bibliographic database. The Islands interface is compared to the Dialog Interface on a search of the INSPEC database.

Additional Comments
The Islands system attempts to graphically display the combinations of search sets in an information retrieval context. While the illustration of the combined searches is helpful, the additional information conveyed by the Islands system may unnecessarily complicate the display. The 3D features of this interface are not well justified in the explanatory article, although the graphical effects of Boolean operators is quite effective.


TextScape

TextScape

Rossi, A.M.; Varga, M.(1999). Visualisation of massive retrieved newsfeeds in interactive 3D. In E. Banissi, et al. (Eds.), Proceedings of the International Conference on Information Visualization, 1999. (pp 12-17).


.pdf available from the IEEE Digital Library

Author's Abstract
This report describes work carried out under the Automatic Information Retrieval project at DERA, in building a 4D landscape (a TextScape) which acts as a graphical user interface (GUI) to the DERA-OKAPI search engine. Traditionally search engines present a 2D visual interface (usually a windows-icons-menus-pointer (WIMP) interface) to the user. Key words are entered and the 'hits' are displayed usually in a list form. Unfortunately, the number of documents retrieved is often overwhelming and a significant amount of time is still required to read abstracts and sections of documents in order to determine if they are relevant, even after the search engine has performed relevance ranking. The challenge is how to improve on this. We propose to render the information most immediately required onto the visual attributes of a 3D landscape. Shape, colour and size can then project the information that is sought, and mining the object hierarchy visually returns information more rapidly.


Main | Overviews | Classics | Demo Potential | Other systems | User-testing | Bibliographies
Compiled and annotated by Elizabeth Staley for Michael Twidale for Independent Study
Graduate School of Library and Information Science
University of Illinois
501 E. Daniel Street
Champaign, IL 61820
Please send comments or suggestions to e-staley@alexia.lis.uiuc.edu
Last updated 12 June 2000