Theme Search and Sensemaking
Location Snow Mountain Ranch, Fraser, Colorado
Date January 31 - February 4, 2007
Paper Presentations
Ancient Wisdom and Modern Technology: The Search and Sense-making Strategies of The Walking People
John Thomas (IBM Research)
Discussant: Peter Pirolli
+ Abstract - Abstract
The Walking People is the millennia-long oral history of a branch of the Iroquois translated into written English. During their long journey (from Asia to and across North America to the Atlantic and back to the Great Lakes), the people encountered a continuing set of environmental and social challenges. The Walking People continually work to improve their strategies of search and sense-making. We first outline these strategies and enumerate implications for technological support.
Augmented Social Cognition: Understanding Social Foraging and Social Sensemaking
Ed Chi (PARC)
Discussant: Jaime Teevan
+ Abstract - Abstract
We summarized initial results from PARC's Augmented Social Cognition research program. We are interested in understanding the emerging behavior in social software, particularly how they enable social foraging and social sensemaking. We study both Wikipedia and to characterize their evolution.
Capturing Emergent Consensus across Diverse Opinions
Chaomei Chen (Drexel University)
Discussant: Mark Ackerman
+ Abstract - Abstract
Understanding the essence of a large number of diverse opinions is important and challenging to the study of online communities and other emergent social phenomena. We propose a unifying framework and present examples of capturing and differentiating emergent consensus across thousands of conflicting opinions.
CoLiDeS and SNIF-ACT: Complementary Models for Searching and Sensemaking on the Web
Marilyn Blackmon (University of Colorado)
Discussant: Bonnie John
+ Abstract - Abstract
CoLiDeS and SNIF-ACT are empirically-validated, complementary models. SNIF-ACT applies rational analyses to information foraging anywhere on the Web. CoLiDeS describes how people attend to and comprehend information patches on individual webpages. Integrating CoLiDeS and SNIF-ACT would better predict how people forage the Web for information to solve everyday ill-structured problems.
Collaborative search and sensemaking in CiteSeer
John Carroll (Penn State)
Discussant: Dan Russell
+ Abstract - Abstract
We discuss search and sensemaking in the context of the Next Generation CiteSeer project. CiteSeer is a digital repository of research literature in the computer and information science domain, consisting of a specialized search engine and digital library. We are investigating opportunities and mechanisms to enhance it as a collaboratory. We will discuss requirements gathering and analysis, design concepts, and experiments with prototypes, focusing particularly on RSS feeds as notification systems for awareness and tagging as social bookmarking features.
Leverage Points and Tools for Aiding Intelligence Analysts
Stu Card (PARC)
Discussant: John Stasko
+ Abstract - Abstract
Intelligence analysis brings together massive information on the source side with massive knowl-edge on the analyst side under severe task demands and time constraints. As part of a human- and task-centered research and design project, we used a cognitive task analysis of several analysts' to get a nominal characterization of the overall process. We then identified leverage points where construct-ing aids would be likely to be beneficial. We prototyped aids in the areas identified and engaged in technology transfer
Making Sense of Information Scatter on the Web
Suresh Bhavnani (School of Information)
Discussant: Tom Landauer
+ Abstract - Abstract
Recent studies suggest that users often retrieve incomplete healthcare information because of the complex and skewed distribution of facts across relevant webpages. To understand the regularities underlying such skewed distributions, this paper presents the results of three analyses. (1) A distribution analysis describes how facts related to healthcare topics are scattered across high-quality healthcare pages. (2) A cluster analysis of the same data suggests that the skewed distribution can be explained by the existence of three page profiles that vary in information density, each of which play an important role in providing comprehensive information of a topic. (3) A network analysis reveals regularities of how particular facts co-occur in pages. The regularities underlying scatter revealed by the three analyses provide implications for the design of search systems and websites to help users find and make sense of scattered information, and implications for a model to explain the process through which information scatters occurs.
Patterns of Tag Usage across Four Diverse Enterprise Tagging Services
Michael Muller (IBM Research)
Discussant: Robin Jeffries
+ Abstract - Abstract
We compare four tagging-based enterprise services, that respectively stored bookmarks to webpages and documents, to people, to blog entries, and to hierarchically-structured activity records. Analysis of user data and tag data showed diffuse patterns of commonalities and differences across the services. These results will help us to understand emergent work practices, and the value of tagging services to enterprises.
Representational Change in Sensemaking
George Furnas (Univ of Mich)
+ Abstract - Abstract
Sensemaking has been described by Russell et al as a process of representational development, wherein people seek increasingly effective representations to support the task they face. This paper presents two partial models of such representational change, along with discussions of the strengths and weakness of each.
Search and Sensemaking in Design Reuse
Scott McCrickard (Virginia Tech)
Discussant: Scott Klemmer
+ Abstract - Abstract
Reuse of HCI design knowledge could improve design by leveraging previous research. With selection and integration core characteristics of reuse, search and sensemaking play large roles. Reuse of claims--reusable HCI design knowledge components--through a claims library, claims map, and claim relationships can support search and sensemaking.
Search and Sensibility: Tools, Patterns, and Practice in Everyday Search and Sensemaking
Lilly Irani (Google)
Discussant: Clayton Lewis
+ Abstract - Abstract
We investigate the notion of 'everyday sensemaking' using Google Notebook, a tool designed to support clipping, organization, and retrieval within the web-browser. We examine a large number of notebooks and observe search, collection and organizing behavior as part of real-world information gathering tasks, linking users' notebook use to their search practices to their representations of the problem to be made sense of.
Supporting social search with social bookmarking
David Millen (IBM Research)
Discussant: Terry Roberts
+ Abstract - Abstract
In this paper, we explore various aspects of social search supported by a social bookmarking service. These bookmarking services hold great potential to powerfully combine personal tagging of information sources with interactive browsing, resulting in better social navigation. We examine use when deployed in a large organization, through logfile analysis and end-user interviews, to understand how social search is supported. We conclude that social search is supported well through a combination of better personal bookmark management, enhanced social navigation and application-specific search.
Towards a Model of Information Scatter: Implications for Search and Design
Suresh Bhavnani (School of Information)
+ Abstract - Abstract
Recent studies suggest that users often retrieve incomplete healthcare information because of the complex and skewed distribution of facts across relevant webpages. To understand the causes for such skewed distributions, this paper presents the results of two analyses: (1) A distribution analysis discusses how facts related to healthcare topics are scattered across high-quality healthcare pages. (2) A cluster analysis of the same data suggests that the skewed distribution can be explained by the existence of three page profiles that vary in information density, each of which play in important role in providing comprehensive information of a topic. The above analyses provide clues towards a model of information scatter which describes how the design decisions by individual webpage authors could collectively lead to the scatter of information as observed in the data. The analyses also suggest implications for the design of websites, search algorithms, and search interfaces to help users find comprehensive information about a topic.
Video Activity Tracking: Supporting Sensemaking with Geographic Cues
Andreas Girgensohn (FX Palo Alto Laboratory)
+ Abstract - Abstract
We present designs that provide geographic cues to aid cross-camera tracking and a study thats compares user performance and preferences against a traditional interface. Our spatial video player orients nearby video feeds with the main camera view. We discuss implications for the design of video surveillance interfaces and multi-video applications.
What are you looking for? An eye-tracking study of information usage in Web search
Ed Cutrell (Microsoft Research)
Discussant: Peter Pirolli
+ Abstract - Abstract
We describe a study that used eye tracking methodologies to explore the effects of changes in the presentation of search results. We found that adding information to the contextual snippet significantly improved performance for informational tasks but degraded performance for navigational tasks. We discuss possible reasons for this difference and the design implications for the better presentation of search results.
Boaster Presentations
Evaluating Ripple: Experiences from a Cross Pollinated SE-UE Study
Pardha Pyla (Virginia Tech)
+ Abstract - Abstract
This paper describes the preliminary findings from an evaluation of the Ripple framework in an educational setting. This framework provides a development infrastructure to foster communication between software engineers and usability engineers thereby connecting usability and software engineering life cycles in a cooperative and complementary environment.
Gender Differences in Remote Trust and Performance with Initial Social Activities in CMC Environment
xiaoning sun (Drexel University)
+ Abstract - Abstract
This proposed empirical study aims to investigate the impacts of both communication media and initial social activities on trust development and performance for different gender composition groups in a virtual environment. The study may provide insights into ways of improving performance of diverse group compositions in real world virtual collaborations.
Scaffolding the construction of scientific explanations from multiple online sources
Sebastian de la Chica (University of Colorado at Boulder)
+ Abstract - Abstract
Writing scientific explanations presents unique pedagogical opportunities for learners to engage in a realistic and educationally relevant science task. Educational summaries generated from digital library collections and encoded as concept maps provide the basis for educational scaffold design to support learners writing scientific explanations from multiple online sources.
Supporting creativity: Investigating the role of computer-supported awareness in distributed collaboration
Umer Farooq (Penn State)
+ Abstract - Abstract
The role of computer-supported awareness in supporting creativity as a long-term, collaborative activity has not been investigated. In my dissertation, I am investigating the design and evaluation of activity awareness features to support creativity in small groups working in a distributed setting on a collaborative scientific activity over multiple weeks.
Web Accessibility for People with Cognitive Disabilities
clayton lewis (university of colorado)
+ Abstract - Abstract
A placeholder with null paper. I will be advising a body on strengthening regulation of accessibility for people with cognitive disabilities. A worry is specifying objective tests for things like comprehensibility of text. I'd love to discuss this with the insightful attendees of HCIC (the rest of you, forget it.)
Towards a Tool for Predicting User Exploration
Leonghwee Teo (HCI Institute, CMU)
+ Abstract - Abstract
CogTool-Explorer is a tool to predict user exploration choices given a user interface and task. We describe the integration of components that make up CogTool-Explorer, and how it interprets the text of both the task description and interface elements to make its exploration choices. We present CogTool-Explorer's performance on a web navigational search task compared to human performance.
Making Sense of Community Expertise Networks
jun zhang (School of Information, University of Michigan)
+ Abstract - Abstract
Seeking expertise from other people is one of the most common ways for people to solve real life problems. This paper describes our research on social networks constructed from people's expertise sharing activities in online communities. A concept -- Community Expertise Networks (CEN) -- is proposed and discussed. (The attached paper is submitted to WWW2007)
Exploring and investigating: Supporting high-level search activities
Gina Venolia (Microsoft Research)
+ Abstract - Abstract
Much work has been done to improve search as an isolated act, yet little has been done to understand search as it relates to higher-level patterns of behavior, or to develop user interfaces to support these patterns. In this paper, we analyze exploratory and investigative search processes that involve performing several related searches over a period of time. We then discuss requirements for interfaces to support these tasks, and describe three prototype systems.
ResultMaps: Search Result Visualization for Hierarchical Information Spaces
Edward Clarkson (Georgia Tech)
+ Abstract - Abstract
We have developed ResultMaps, a treemap-based visualization we use in two digital library search facilities. A ResultMap provides an overview of the entire document space, and facilitates cluster and outlier detection. We suggest an evaluation plan for studying ResultsMaps, and the potential of applying them to faceted browsing environments.
aIRPLane: An Information Retrieval Pattern Language
Christine Wania (Drexel University)
+ Abstract - Abstract
Interaction patterns and pattern languages have been discussed for years in HCI literature yet there have been few empirical studies conducted. In this paper we describe aIRPLane: An Information Retrieval Pattern Language, its discovery, and the experimental design we use to examine its impact on the design of interfaces.
Sensemaking Handoff: Theory and Recommendations
Nikhil Sharma (UM)
+ Abstract - Abstract
Sensemaking work is often handed off between people. Yet this handoff can cause problems, somewhat similar to an 'interruption'. This dissertation research examines the issues related to sensemaking handoff by integrating existing theories and drawing predictions about the effects of premature handoff. These predictions and the related design recommendations for systems will be verified using short-term ethnography, interviews, laboratory observations and experiments.
What Information Do Developers Need?
John Daughtry (The Pennsylvania State University)
+ Abstract - Abstract
Software development is a complex process of search and sensemaking. We are interested in this activity at the API level of analysis. This work can inform search processes (e.g. ' methodologies) and tools (e.g. ' code search engines). In this paper I discuss our research in this area.
Range: Exploring Proxemics in Collaborative Whiteboard Interaction
Wendy Ju (Stanford University)
+ Abstract - Abstract
Range is an interactive whiteboard designed to support collocated, ad-hoc meetings. It employs proximity sensing to proactively transition between ambient and authoring modes, clear space for writing, and cluster ink strokes. The inspiration for Range stems from longitudinal studies of student design teams, in which we observed that shifts in collaborative activity correlated with changes in the users' physical proximity to the whiteboard.
Knowledge Transparency in Collaborative Software Development
Jan Chong (Stanford University)
+ Abstract - Abstract
This is an ethnographic study of how programmers on two software development teams gather and share knowledge in the course of work. I investigate how these activities are affected by the configuration of their work practices and environment as well as the implications for the development of knowledge management tools.
Supporting Data-Based Decision-Making for Caregivers through Embedded Capture and Access
Julie Kientz (Georgia Institute of Technology)
+ Abstract - Abstract
In continual care settings, caregivers must make frequent decisions to ensure optimal care. For my dissertation, I am exploring the use of embedded capture and access to support decision-making for caregivers. I am applying this in two domains: therapy for children with autism and recording-keeping for parents of newborn children.
Storytelling and Sensemaking: Eliciting the Stories Embedded in Personal Digital Photographs
Brian Landry (Georgia Institute of Technology)
+ Abstract - Abstract
Digital photographs capture moments in time and are typically part of much larger experiences. Storytelling is often used to make sense of and convey these experiences. We are proposing a story-based model for annotation and search to support sensemaking and storytelling of personal experiences.
Virtual Community Maintenance with a Collaborative Repository
Derek Hansen (University of Michigan)
+ Abstract - Abstract
This paper presents the results of a case study examining how a wiki repository is used to help overcome some of the community maintenance challenges common to help-based email list discussions. These include keeeping the discussion on topic, avoiding "holy wars", welcoming newcomers, and promoting participation.