J. Blustein's Publications

Colour scheme: light | dark | None

Short List

Titles are links to detailed list (below).

Refereed

Theses and Dissertations

Non-Refereed

Significant Electronic Publications

Titles are links to the documents.

Other

Author's Address

Detailed List

This detailed list is in chronological order, from most recent to earliest.  Articles tagged with {H*} are hypertexts.  Articles tagged with [R] were refereed.

Menu

[R] Impact of Spatial Visualization Aptitude on WWW Navigation
James Blustein, Ishtiaq Ahmed, Haris Parvaiz, Ching-Lung Fu, Caixia Wang, Alexander (Sandy) Chapman, and Yeming Hu
The Ergonomics Open Journal Vol. 2 pp. 80-87
DOI: 10.2174/1875934300902020080
Abstract

Although the underlying mechanism is not well understood, there is considerable evidence that the constellation of cognitive factors known as spatial ability influences users' performance in information spaces. Evidence of the effect in the computer science literature is contradictory: some studies show that techniques, which support users with lower aptitude, retard performance by those with higher aptitude. We have investigated the effect of the visualization subfactor in a real-world navigation task using location menu breadcrumbs and Dillon's IMRD task.

We compared the navigational styles and success rates in an answer seeking task using both standard and menu breadcrumbs in a large website. The higher aptitude group was significantly more efficient and used the Back button less than the lower aptitude group.

We discuss implications for explaining why spatial aptitude affects success with hypertext, the potential for practical application, and ongoing follow-up work.

Available from publisher's website:
[R] Electronic Commerce On-Site Search Services: A state of the art review
Nan Li, Peter Hitchcock, James Blustein and Michael Bliemel
WEB 09: The Eighth Workshop on e-Business
15 December 2009
Phoenix, Arizona, USA
[R] Showing Sharing: Information displays for managing shared files
Tara Whalen, Elaine G. Toms, and James Blustein
Symposium on Computer Human Interaction for Management of Information Technology (CHIMIT) 08
14-15 November 2008
San Diego, USA
Abstract

Within the workplace setting, people need to provide sufficient access to files to allow collaboration, but not so much that sensitive files are inadvertently exposed. Yet evidence suggests that file sharing problems decrease security and interfere with collaboration. A potential solution for managing these problems is to present the user with clear information about file sharing settings and activities. Current file managers either hide the information or simply do not provide it. Using an awareness framework, we identified the core information that users need to be aware of in file sharing situations, performed two studies to determine how to best represent those concept as labels and icons, and developed a prototype for a file manager that reveals file sharing activity. The results of this design activity can be adopted for other file sharing applications, improving their security and collaborative usability.

[R] Natural Search Pointers — a query formulation method for structured information search
Marek Lipczak, James Blustein, and Evangelos Milios
2008 Conference on Human System Interaction (HIS'08)
25-27 May 2008
Krakow, Poland
DOI: 10.1109/HSI.2008.4581436
Abstract

Despite a wide variety of new solutions, structured information search still has only one practical approach — form-based interface. A key limitation of this interface is poor handling of iterative search. While browsing the results users have to memorize all new search constraints, go back to the form, and enter them into the appropriate fields. To overcome this obstacle we created Natural Search Pointers — a structured information search interface, which formulates search queries based on information highlighted by a user while browsing the search results. NSP can be used as an extension of any standard form-based interface for consumer-oriented database search engines. Comparison of traditional form-based interface and its NSP extension shows that in iterative search tasks NSP makes finding information faster and more convenient.

Availability:
[R] File Sharing and Group Information Management
Tara Whalen, Elaine Toms, and James Blustein
Personal Information Management: PIM 2008 workshop at CHI conference
5-6 April 2008
Florence, Italy
Abstract

Groups that interact digitally often communicate and collaborate through shared files. Managing access to these shared files can be complex: one needs to provide sufficient access to allow collaboration, but not so much that sensitive files are inadvertently exposed. Preliminary evidence suggests that file sharing problems decrease security and interfere with collaboration. We collected data in order to develop a comprehensive picture of how users carry out file sharing, and the problems they encounter. We present our results, and describe the implications for group information management. We also describe some design ideas, and discuss how these can be adapted for file management in the disappearing desktop context.

Availability:
[R] Impact of Spatial Visualization Aptitude on WWW Navigation
James Blustein, Ishtiaq Ahmed, Jason Satel, Haris Parvaiz, Ching-Lung Fu Yeming Hu, and Caxia Wang
Workshop on Cogntion and the Web: Information Processing and Learning
pp. 185-192
24-26 April 2008
Granada, Spain
Availability:
[R] Influence of Spatial Ability in Navigation
Ishtiaq Ahmed and James Blustein
International Journal Of Web Based Communities Vol. 2, Issue 2, pp. 183-196 (2006)
DOI: 10.1504/IJWBC.2006.010309
Draft Abstract

Spatial implications of the commonly used navigation metaphor have lead many researchers to investigate the relation between individual differences and navigation. This study presents an exploratory survey on the influence of spatial ability, the most incisive aspect of individual difference for navigation, when people try to accomplish their goal in the information space. There are still different opinions about the potential of visual mediators for people with low spatial ability that help in refraining from the state of being disoriented or lost in the huge information content. The findings motivated us to design a research framework focusing a special visual look-ahead breadcrumb tool. Breadcrumbs are a type of navigational aid that help users from being lost in large websites by providing information about their location in the site hierarchy. The framework is presented after the preliminary discussion. The results from the experiment report that spatial ability influenced navigation efficiency on the Web. The effect of spatial ability was marginally significant on the use of Back button during navigation. However, there was no interaction effect found between spatial ability and the visual look-ahead breadcrumbs on the navigation performance of the users.

Availability:
  1. Official: record at the International Journal of Web Based Communities
  2. Author's copy
  3. Draft available as a tech. rep.: TR-CS-2005-12
[R] An Evaluation of Menu Breadcrumbs for the WWW
James Blustein, Ishtiaq Ahmed, and Keith Instone
ACM Hypertext Conference 2005
pp. 202 – 204
Availability
  1. Pre-printed as a tech. rep.: TR-CS-2005-11
  2. from ACM DL with DOI 1083356.1083394
Information Visualization for an Intrusion Detection System [a poster]
Ching-Lung Fu, James Blustein, and Daniel L. Silver
ACM Hypertext Conference 2005
Abstract available as
  1. Pre-printed as a tech. rep.: TR-CS-2005-10
  2. from ACM DL with DOI 10.1145/1083356.1083419
PDF version of the poster available as:
[R] Information Visualization for Intrusion Detection
James Blustein, Daniel L. Silver, and Ching-Lung Fu
PST '05
Abstract

Spatial hypertext was developed from studies of how humans deal with information overflow particularly in situations where data needed to be interpreted quickly. Intrusion detection requires security managers of large networks to rapidly respond (often in real-time) to masses of information. Users of such systems need to recognize large developing patterns in masses of data, they prefer to work individually (although they must function in collaborative groups), and they rely on their intuitions more than deductive logic. Such users have particular personality characteristics and job needs which can be well served by interfaces which use a spatial hypertext model. Also, like most users, they prefer to be in charge of the process that they use the computer as a tool to assist with. The architecture proposed in this article is based on spatial hypertext and machine learning. That interface design allows for a great deal of interface flexibility and user control. The article discusses in detail how spatial hypertext, and the proposed architecture in particular, can well fulfill the needs of intrusion detection system users through personalized information filtering.

Availability:
[R] Chronological Sampling for Email Filtering
Ching-Lung Fu, Daniel Silver, and James Blustein
Proceedings of the Workshop on Machine Learning for User Modeling: Challenges at UM '05 (10th Int'l Conf. on User Modeling)
(2005)
Abstract

User models for email filtering should be developed from appropriate training and test sets. A k-fold cross-validation is commonly presented in the literature as a method of mixing old and new messages to produce these data sets. We show that this results in overly optimistic estimates of the email filter's accuracy in classifying future messages because the test set has a higher probability of containing messages that are similar to those in the training set. We propose the k-fold chronological cross-validation method that preserves the chronology of the email messages in the test set.

Availability:
[R] Navigation in Information Space: How Does Spatial Ability Play A Part?
Ishtiaq Ahmed and James Blustein
In Proc. the Second Int'l Conf. on Web Based Communities (WBC '05)
pp. 119 – 125
Feb. 2005, Algarve, Portugal
(Conference sponsored by IADIS)
Availability:
  1. Definitive version available from the IADIS DL
  2. Author's copy
[R] Navigation in Information Space
Ishtiaq Ahmed and James Blustein
In Proc. the Second Int'l Conf. on Web Based Communities (WBC '05)
pp. 281 – 286
Feb. 2005, Algarve, Portugal
(Conference sponsored by IADIS)
Availability:
  1. Definitive version available from the IADIS DL
  2. Author's copy
[R] Personal Glossaries on the WWW: An Exploratory Study
and
{H*} Personal Glossaries on the WWW: An Exploratory Study (Hypertext)
James Blustein and Mona Noor
ACM Symposium on Document Engineering 2004
Abstracts:
  1. Traditional Form

    We examine basic issues of glossary tools as part of a suite of annotational tools to help users make meaning from documents from unfamiliar realms of discourse. We specifically evaluated the performance of glossary tools for reading medical information about common diseases by users with no formal medical education.

    We developed both automatic and an editable glossary tools. Both of them extracted definitions from the text of articles. Only the editable glossary tool allowed users to add, delete, and change entries.

    Both tools were evaluated to find out how useful they were to users reading technical articles online. The analytical results showed that user performance improved without increasing total reading time. The glossary tools were effective and pleasing to users at no decrease in efficiency. This experiment points the way for longer-term studies with adaptable tools, particularly to help users unfamiliar with technical documents. We also discuss the rôle of glossaries as part of a suite of annotational tools to help users make personal (and therefore meaningful) hypertextual document collections.

  2. Hypertext form

    This article presents an experiment with interfaces for glossaries attached to electronic documents.

    The authors argue that glossaries are a distinguished form of annotation in need of study, and suggest relevant classifications of such tools. In particular they distinguish between (a) glossaries that are tied to single documents or those that can be used with many documents, and (b) glossaries for use only by their authors (so-called personal glossaries), and those for general use (termed shared glossaries). Unlike annotation in general, glossaries have a structure composed of terms, and definitions (which may also include references to other related materials).

    Using lessons garnered from earlier work on usability issues, the researchers concentrated on users' needs and the potential for glossaries to alter the ways people use electronic documents. An automatic glossary tool and an editable glossary tool were developed. Both tools used the same definitions extracted from the text of articles. Unlike the automatic tool, the editable tool allowed users to alter, delete and add new entries. Both tools were evaluated to find out how useful they were to users reading technical articles online. Measures of speed, comprehension, and satisfaction were collected. Forty users were given two online articles about selected health conditions to read in a fully-randomized mixed design with each particpant completing two sessions (one with and one without a glossary).

    Although they accessed glossary entries regardless of which tool they were given, users did not employ the specific features of the editable glossary. However, users found both of the tools useful and easy to use. Furthermore, users commented that if they knew that they could use the glossary after the experimental session, or that they would need to understand the terms for a test or in their job, then they would have used the editable glossary more.

    The analytical results showed that user performance improved without increasing total reading time. The glossary tools were effective and pleasing to users at no decrease in efficiency. The experiment points the way for longer-term studies with adaptable tools, particularly to help users unfamiliar with technical documents.

    An additional feature of this article is its formatting as a hypertext. The design of the hypertext is based on studies of how people process written technical materials and is intended as a paradigm for future publications.

Availability
  1. Official: DOI:10.1145/1030397.1030409
    • Hypertext article available as from the Html link
    • Traditional article available as a PDF™ file
  2. Preprints:
    1. preprint of hypertext article
      • available in XHTML (22 nodes, 60 files)
      • available in tar.gz format (approx. 1.2 Mb)
    2. preprint of traditional article (in PDF™, 3 letter-sized pages)
Related technical report: CS-2003-09
  • Online Glossary Tools for Technical Reading by Mona M. Noor, Dec. 2003
[R] {H*} A Personal Information & Knowledge Infrastructure Integrator (2004)
K. Andrew Edmonds, James Blustein, and Don Turnbull
Journal of Digital Information (JoDI)
Vol. 5, Issue 1, Article number 243
Availability: link to article at journal website
Abstract

The Next Big Thing is being grown organically, cultivated by software developers and pruned by personal weblog publishers. The rising weblogging space of the Internet is looking more like traditional hypertext than the World Wide Web of the 1990s. The ways in which weblogging has evolved beyond the previous limitations of the Web as hypertext will be explored and the ways weblogging is evolving towards common-use hypertext destined to play a critical role in everyday life. We have a vision of a universal information management system built on extending the traditional hypertext framework. In our utopian future, everyone will use tools descended from today's blogs to structure, search, and share personal information as well as to participate in shared discussion.

We begin by expressing a vision of common-use hypertext for information management and interpersonal communication. This vision is grounded in the rapid evolution of weblogs and known issues in information systems and hypertext. The practical implications of who will use these systems, and how, is expanded into a detailed exploration of weblogs now and in the future as usage scenarios. After recapping the current issues facing the weblogging community we look to the long-range implementation issues with optimism.

Our system is forward-looking yet realistic. The activities the system will support are extrapolated from recent developments in the online community and most of the sketches of implementation are based on current approaches. It is of more than passing interest that the features we extrapolate were all described by Ted Nelson as early hypertext ideals. Of particular interest is that the features are now being implemented because of perceived immediate need by communities of interest.

[R] Improving Intrusion Detection Systems Through Heuristic Evaluation (2004)
Andrew T. Zhou, James Blustein, and Nur Zincir-Heywood
In 17th Annual Canadian Conf. on Electrical and Computer Engineering
May 2004 (vol. 3, pp. 1641 – 1644)
INSPEC Accession Number: 8082900
Availability:
Abstract

This work is a report on efforts to improve the usability of intrusion detection systems. Specifically, we first conducted a worldwide survey of system administrators from different countries and economic sectors to understand the state of practice in security management with a particular focus on intrusion detection systems (IDSs). Then, based on these survey results and in depth interviews, we developed new heuristics to measure the effectiveness and efficiency of IDSs. The comparison of our refined heuristics and Nielsen's general heuristics on Snort, Snortsnarf and our proposed interface show that evaluators using our heuristics find significantly (p<0.0002) more of the problems. Also, evaluations with both sets find fewer problems in our interface than in Snort or Snortsnarf.

Related technical report: CS-2003-06
  • The State of Network Security Management: Issues and Directions by Andrew T. Zhou et al., May 2003
Hierarchical Task Analysis for Intrusion Detection Systems (2004)
Shibu Bashir
A research paper submitted to the Faculty of Computer Science [at Dalhousie University] in partial fulfillment of the requirments for the degree of Masters of Electronic Commerce
Availability:
Spatial Ability and Information Shape: When do individual differences matter
James Blustein and Jason Satel
In Third Workshop on Spatial Hypertext, 2003.
Abstract

Psychologists have long known that cognitive differences between individuals can significantly affect performance on a variety of tasks. Several recent studies have shown that so-called spatial reasoning ability has a significant effect on users success with (spatial and non-spatial) hypertexts. If we understood why spatial ability has such a strong effect on success with hypertext then we could adapt hypertext for use by different types of people, and explore new types of presentation.

It is clear that the different success rates are not solely due to the hypertext systems and the need to integrate two-dimensional spatial data (as with spatial HT systems like VKB) but rather the need to make sense of, and navigate in, multi-dimensional structures of meaning. A. Dillon and D. Schaap refer to some of these issues as information shape.

Despite clear evidence that spatial reasoning ability affects success with hypertext, studies of the effect are remarkably vague about what subfactors were assessed. We have made a preliminary analysis of these studies to determine their common components. We present both a survey of studies of hypertext that have found such effects, and an investigation into the underlying causes.

Available as Dalhousie Computer Science Technical Report CS-2003-11
Hypertext 2002: Proceedings of the Thirteenth ACM Conf. on Hypertext and Hypermedia
Kenneth M. Anderson, Stuart Moulthrop, and James Blustein (eds.)
Published by The Association for Computing Machinery, ISBN 1-58113-477-0.
[R] Methods of Generating and Evaluating Hypertext
James Blustein and Mark S. Staveley
In Annual Review of Information Science and Technology (ARIST) vol. 35 2001
Martha E. Williams, ed.
Section 6 (pp. 299 – 335)
Published on behalf of the American Society for Information Science and Technology by Information Today, Inc.; 2001.
[R] Automatically Generated Hypertext Versions of Scholarly Articles and Their Evaluation
James Blustein
In ACM Hypertext 2000 Conf.
May – June 2000

Available in PDF™ (approx. 575 000 octets):

Abstract

The overall objective of this work is to develop and evaluate ways of automatically incorporating hypertext links into pre-existing scientific articles. Some readers like hypertext even when it is not as useful to them as the linear document from which it was generated. Hypertexts must therefore be evaluated for usefulness and acceptability. We describe rules for making links and an experiment using two methods of applying those rules, to show how such rules should be evaluated, and to see if they truly help people. In addition to measures of performance we also collected measures of preference. The effectiveness of these links was evaluated by testing with people. Performance was determined by measuring the accuracy and inclusiveness of answers to questions about the article, and written summaries. Readers judged the quality of links (and thereby the quality of the rules used to forge them) and the overall effectiveness of the hypertext. Most readers did not read the entire articles in the time allotted. Readers had no preference for articles with or without novel link types, but they did have a strong preference for definition and structural links over (novel) semantic links. Readers of documents with only structural links had comprehension scores that were inversely proportional to their satisfaction ratings. No performance difference was detected.

Dalhousie Computer Science Technical Report CS-2002-06: Recording Links Followed in a Single HTML Document
James Blustein
(Formerly in press at The Perl Journal, but retroactively rejected by new management)
Availability:
Abstract

This report describes in detail a combination of tools (written in Perl, Javascript and the hypertext markup language (HTML) that were used to record the links users followed within a single web page.

The method was developed specifically for use with the Perl 5 language and the Netscape World Wide Browser (version 4 for Unix) but can be easily adapted to other browsers. The main drawback to the method is that users must click the back button twice to return to the previous link. It should be a simple matter to eliminate that drawback using today's browsers.

Panel: Adaptive Hypermedia
James Blustein and Luis Francisco-Revilla
sigWEB Newsletter vol. 8, no. 2, June 1999
Available (in PDF™) from ACM
Hypertext Versions of Journal Articles: Computer-aided linking and realistic human-based evaluation
William James Blustein
Doctoral thesis
April 1999
Department of Computer Science, University of Western Ontario
London, Ontario, Canada
A PostScript version (encoded by gzip) is available for FTP from ftp.csd.uwo.ca in the pub/thesis directory.
Abstract

My overall objective is to develop and evaluate ways of automatically incorporating hypertext links into pre-existing scholarly journal articles. I describe a rule-based approach for making three types of links (structural, definition, and semantic). Structural links are a way of making explicit some connections between parts of the text. Definition links connect the use of a term, defined elsewhere in the document, to that definition. Links that connect parts of text that discuss similar things are semantic links. I distinguish several types of semantic links.

I use two information retrieval (IR) systems (Cornell's SMART system and Bellcore's Latent Semantic Indexing) to select links based on the content of the articles. I conducted an experiment to compare the performance of the links forged using these two systems.

The effectiveness of the links (and the rules used to make them) is tested by people reading the hypertext versions for information under a time constraint. A within-subjects experimental design was used. Each of the nineteen experimental participants read one version of each of three scholarly articles in a different hypertext form (one had only simple links, the others had definition links and semantic links selected using one of the IR systems). Subjects' preferences were also measured.

Although I used three survey articles from published sources for my evaluation experiment there was no difference in reader preference or performance on the basis of article. Subjects ratings of the utility of the various links shows a significant preference for structural links over semantic links. Definition links were preferred to structural links, although the result was not significant. No difference between the links created using the two IR systems was detected.

However there were significant differences in the times that readers spent on documents created using the various treatments When they read in documents with only structural links readers were more likely to have read the whole article, and their satisfaction scores were inversely proportional to their comprehension score.

The method of evaluating hypertext versions of journal articles for use by researchers may be applied to other hypertext versions.

[R] {H*}Evaluating Automatically Generated Hypertext Versions of Scholarly Articles
James Blustein
Hyped-Media to Hyper-Media: Toward Theoretical Foundations of Design, Use and Evaluation
Workshop at CHI '98 conference
Organized by N. Hari Narayanan
19 – 20 April 1998
ACM SIGCHI
Los Angeles, California
Note: The talk was based on an early version of my PhD research. My PhD proposal in brief and completed PhD thesis are also available.
Availability: link to article
Abstract

In this paper I present an experimental approach to the evaluation of a type of hypermedia application. My overall objective is to develop and evaluate ways of automatically incorporating hypermedia links into pre-existing scholarly journal articles. The focus of this paper is the evaluation method. My method allows the results to be applied to other documents than just those tested.

To properly convert ordinary documents into useful hypermedia two constraints must be satisfied: the links must be useful to the readers and the risk of disorientation introduced by the new structure imposed by the links must be minimized. I describe a rule-based approach for making links. In my experiment I use two methods to detect when the rules should be applied. The effectiveness of the links is tested by people performing realistic tasks. Readers judge the quality of links (and thereby the quality of the rules used to forge them) and the overall effectiveness of the hypermedia.

[R] A Design for the Construction and Evaluation of an Automatic Hypertext Generator
James Blustein
Communication and Information In Context: Society, Technology, and the Professions
Proceedings of the 25th Annual Conf./Travaux du 25e congrès annuel
Canadian Association for Information Science (CAIS)/Association Canadienne des Sciences de L'information (ACSI)
Edited by Bernd Frohmann
8 – 10 June 1997
Learned Societies Congress/Congrès des sociétés savantes
Memorial University of Newfoundland
St. John's, Newfoundland
Note: The talk was based on my PhD proposal. The proposal in brief is also available.
Abstract

My overall objective is to develop and evaluate ways of automatically incorporating hypertext links into pre-existing scientific articles. Hypertext can support all of the ways we believe people use printed versions of articles and can add additional useful features. However many readers find hypertext confusing and not all texts are suitable for conversion to hypertext. Some readers like hypertext even when it is not as useful to them as the linear document from which it was generated. Hypertexts must therefore be evaluated for usefulness.

To properly convert ordinary documents into useful hypertext two constraints must be satisfied: the links must be useful to the readers and the risk of disorientation introduced by the new structure imposed by the links must be minimized. I describe rules for making links and two methods to detect when the rules should be applied.

I propose to provide links by applying rules based on the content of the articles using two methods: the Cornell's SMART system and Bellcore's Latent Semantic Indexing. I will evaluate the effectiveness of these links by testing with people. Readers will judge the quality of links (and thereby the quality of the rules used to forge them) and the overall effectiveness of the hypertext.

[R] Methods For Evaluating the Quality of Hypertext Links
James Blustein, Robert E. Webber, and Jean Tague-Sutcliffe
Information Processing & Management (March 1997, vol. 33, no. 2, pp.255 – 271)
Abstract

We present two methods for evaluating automatically generated hypertext links. The first method is based on correlations between shortest paths in the hypertext structure and a semantic similarity measure. Experimental results with the first method show the degree to which the hypertext conversion process approximates semantic similarity. The semantic measure is in turn only an approximation of a user's internal model of the corpus. Therefore we propose a second evaluation method based on measuring user's performance using hypertext. Finally, we discuss the advantages and disadvantages of computer versus human evaluation, respectively.

Document record and abstract available
Implementing Bit Vectors in C
James Blustein
Dr. Dobb's Journal (August 1995, vol. 20, issue 8, #233)
Note: An updated version is available at <URL:http://www.csd.uwo.ca/~jamie/BitVectors/>. The code should be available online from the DDJ FTP site (see http://www.ddj.com).

Bit vectors provide an extremely space- and time-efficent means of implementing arrays of Boolean values.

IR-STAT-PAK
James Blustein, Jean Tague-Sutcliffe
Presented at SIGIR '95 Conf. (July 1995)
The C code is available from the Retrieval Group of NIST. A brief (4 page) overview is available as a PostScript file, a PDF™ file, in XHTML 1.0 format and even in HTML 4.0 format.

A program to compute descriptive and analytic statistics for the TREC IR trials.

A Statistical Analysis of the TREC-3 Data
Jean Tague-Sutcliffe, James Blustein
in Text Retrieval Conference
National Institute of Standards and Technology
Gaithersburg, MD, U.S.A.
November 1994
Abstract

A statistical analysis of the TREC-3 data shows that performance differences across queries is greater than performance differences across participant runs. Generally, groups of runs which do not differ significantly at large, sometimes accounting for over half the runs. Correlation among the various performance measures is high.

Both an image-only scan of the hardcopy, and a scan which includes optical character recognition are available.  Both are in PDF.  The size of the file that encodes the image-only scan is about 4.8Mb, the version with OCR is about 6.5Mb in size.
[R] Using LSI to Evaluate the Quality of Hypertext Links
James Blustein and Robert E. Webber
in IR and Automatic Construction of Hypermedia: a research workshop
13 July 1995
Maristella Agosti and James Allan, eds.
ACM SIGIR
Available in two formats;
Abstract

Useful hypertext is constrained by the need for users to be able to find documents about similar topics without extensive navigation. We show how examining the properties of a graph built by a document's hypertext links can be used to evaluate the usefulness of the document. To formally measure the quality of hypertext linking in a corpus, we compare the semantic similarity of pairs of documents with the minimum number of links between their corresponding nodes in an analogous hypertext graph. We use the measure of document-to-document similarity computed using latent semantic indexing as our measure of semantic similarity. Our method has been applied to a corpus composed of Usenet messages.

An Evaluation of Tools for Converting Text to Hypertext
William James Blustein
Masters thesis
December 1993/January 1994
Department of Computer Science, University of Western Ontario
London, Ontario, Canada
Abstract

Methods for automatically converting semi-structured text (Usenet messages) into hypertext form using information retrieval methods were investigated. The methods were evaluated using statistical means to determine which will produce hypertext best suited to browsing and searching. Methods were evaluated by comparing a measure of semantic similarity of all document pairs with the shortest path in a graph formed by hypertext links between those documents.

I have made some minor corrections since publication.

See also
significant electronic publications above

Address

Menu

J. Blustein
Faculty of Computer Science
Dalhousie University
6050 University Avenue
Halifax, Nova Scotia
B3H 1W5
Canada

E-mail: <jamie@cs.dal.ca> or <jamie@acm.org>
Fax: (+1)(902)492-1517
http://www.cs.dal.ca/~jamie/pubs/index.html

This document is copyright by its author, J. Blustein.

This document is written in valid XHTML 1.0; This document makes use of cascading style sheets

The symbol used to tag hypertext articles, {H*}, is a modified version of an image © 2005 by Steve Luker and licensed from iStockPhoto.