CIR paper at SIGIR 2014 conference

I’m very happy to report that our paper on Collaborative Information Retrieval (CIR) is accepted for the ACM SIGIR 2014 conference, to be held in July in Gold Coast, Australia.

Here’s the full citation:
Soulier, L., Shah, C., & Tamine, L. (2014). User-Driven System-Mediated Collaborative Information Retrieval. Proceedings of ACM SIGIR 2014 Conference. Gold Coast, Australia: July 6-10, 2014.

And here’s the abstract:
Most of the previous approaches surrounding collaborative information retrieval (CIR) provide either a user-based mediation, in which the system only supports users’ collaborative activities, or a system-based mediation, in which the system plays an active part in balancing user roles, re-ranking results, and distributing them to optimize overall retrieval performance. In this paper, we propose to combine both of these approaches by a role mining methodology that learns from users’ actions about the retrieval strategy they adapt. This hybrid method aims at showing how users are different and how to use these differences for suggesting roles. The core of the method is expressed as an algorithm that (1) monitors users’ actions in a CIR setting; (2) discovers differences among the collaborators along certain dimensions; and (3) suggests appropriate roles to make the most out of individual skills and optimize IR performance. Our approach is empirically evaluated and relies on two different laboratory studies involving 70 pairs of users. Our experiments show promising results that highlight how role mining could optimize the collaboration within a search session. The contributions of this work include a new algorithm for mining user roles in collaborative IR, an evaluation methodology, and a new approach to improve IR performance with the operationalization of user-driven system-mediated collaboration.

Coagmento in research uses

Coagmento has been used in various research projects, even those that do not call for collaboration. Why? Because Coagmento offers one of the most basic and powerful thing for lab-based studies – log data collection as long as the participants are doing something within their browsers.

We keep seeing examples of such usage from around the world. The most recent one was caught at the CSCW 2014 conference in Baltimore. It was  presented by Ryan Kelly of University of Bath in a paper titled Collaborative Web Search in Context: A Study of Tool Use in Everyday Tasks.

Want to use Coagmento for your research studies? It’s available as an open-source tool from Coagmento website, or you can opt for Coagmento Collaboratory if you’re a bit tech-savvy and want a greater flexibility.


Coagmento Collaboratory

We are happy to announce the availability of Coagmento Collaboratory – an open-source toolkit for constructing user studies with individual or collaborative users.

Coagmento Collaboratory is a version of Coagmento designed to be:

  • Modularized: Coagmento Collaboratory splits up functionality into separate classes which are easier to understand and maintain
  • Extendable: We provide stateless web-services so you can develop your own tools on top of our framework.
  • Documented: Coagmento Collaboratory is well documented through GitHub.

The framework contains the following:

  • Core classes which provide the foundation for Coagmento Collaboratory giving APIs for accessing user data (Bookmarks, Snippets, etc.) and research studies (Projects, Stages)
  • Web-services for accessing the core classes through HTTP requests
  • The Firefox plugin which uses the web-services as an example of practical usage of Coagmento Collaboratory

Find more information and download links from Note that we are still continuing to do more development and welcome contributions from others interested in this project. Comments and feedback always welcome!

Roberto’s dissertation

PhD Dissertation by Roberto I. González-Ibáñez: “A Study of Positive and Negative Affective States in Collaborative Information Seeking” (Defended in Sept. 2013 at Rutgers University; Supervised by Chirag Shah)

Emotions and other affective processes have long been considered essential elements in people’s lives. Whether during intimacy or in social contexts, human beings experience a wide spectrum of emotions every day, all the time. Despite emotion research conducted in various domains, little is known about the role of affects, emotions, feelings, and mood in the information search process, especially when this is carried out by teams. In this regard, this dissertation aimed to understand whether the affective dimension plays a role in collaborative information seeking (CIS) through four research objectives: (1) study how initial affective conditions influence information practices; (2) investigate what affective processes are typically expressed and experienced in information search; (3) examine how initial affective conditions and those derived from social interactions during the collaboration process influence team performance; and (4) study positivity ratio in collaborative search and their relation to team performance. To accomplish these research objectives, a controlled lab study with 135 participants distributed in fixed experimental conditions and a control group was conducted. In each experimental condition, participants were individually treated with affective stimuli in order to elicit positive and negative affective states.

Results from this study suggest that initial affective states may define and/or shape information processing strategies. Additionally, in collaborative settings, it was found that the interplay of similar or different affective processes could change the way searchers interact with each other, their frustration levels, affective load, and the quality of their work. This dissertation and the findings presented have theoretical implications in the study of collaborative and individual information seeking. Specifically, it gives the affective dimension a central role that could define the way people search, evaluate, and make sense of information. In terms of practical implications, if affective processes play such a key role in information seeking, this may redefine the design of information system by incorporating the ability to identify searchers’ initial affective states and provide the necessary resources to support their information processing strategies. Finally, this dissertation also contributes with a research framework and a methodological approach to carry out experimental evaluations to investigate the role of the affective dimension in both collaborative and individual information seeking.


CIS book preface

We live in a society where information is, without a doubt, a powerful force. This statement may sound like a cliche, but it always amazes me how often we forget. May be that’s the purpose (or should be) of the technology that surrounds such information. The other aspect of information technology that amazes me is the fact that it is so new, considering human history. The clock that tells us the precise time of the day dates back only to the sixteenth century. The base 10 numbering system is only 500 years old, and mechanical devices used to calculate and present information have existed for only a couple of centuries. Of course, today when we say “information technology”, we are probably thinking about computers and other digital devices, and they are merely a few decades old.

What intrigues me the most is how we have been able to integrate such new concepts and technologies with long-standing human behavior. Take for example, working in groups and living as a society. This behavior has proven to be extremely important for the survival and prosperity of our species. Back in the days of hunting together to today’s office work, mankind has understood the need to work and thrive together. It is this behavior – the one of collaborating with each other – that has made it possible to achieve great feats in the history. How else can one man (or woman) build the pyramids or crack the human genetics code.

Of course, not all problems call for people working together. While Einstein had help and drew inspirations from others, he did come up with many significant findings himself. Leonardo de Vinci and Picasso, similarly, worked alone. Claude Shannon, considered to be the father of digital information age, was known to have worked in solitude behind closed doors. But let’s put geniuses aside and talk about the remaining 99.99% of us (which of course, still includes a lot of smart people!). We do, often need to work in collaboration. I’m sure even Einstein needed help placing his furniture in his Princeton house; he was a genius, not a superhuman!

This book is about those times when people work in collaboration – an eternal human behavior, in the light of new and innovative technologies in the information age that we live in. More precisely, it is concerned about situations pertaining to information retrieval/seeking/sense-making where people are collaborating or should be collaborating.

One may ask – Why this book? Why now? There is a simple two-fold answer to both these questions. Using technology to understand and support collaborative behavior has been around for a while – what is known as Computer-Supported Cooperative Work (CSCW), but it is in the recent years that we have seen more specialized attention given to applying CSCW methods and frameworks for information seeking situations. On the other hand, the field of Information Retrieval (IR, or broadly speaking IS, information seeking) has found (or re-realized) the importance of considering social and collaborative aspects of search, synthesis, and information use.

This has led to a newly developed interest in the field that is still emerging at the intersection of several other well-established fields, including CSCW, IR, HCI, and social media/networking. This book as an attempt to introduce the relatively young domain of collaborative information seeking (CIS) research by discussing how it came to be, what it currently offers, and where it is headed next. The best part is that we all get to define and contribute to this future.

Personally, my journey on this path started during the summer of 2007 when I was an intern at FXPAL, working with Gene Golovchinsky and Jeremy Pickens. Back then, we worked on something called Collaborative Exploratory Search (CES), and argued that IR systems need to have “smart” components that could mediate collaborative activities and produce results that are “better” than any individual IR process. And we succeeded with at least one kind of situation (time-limited, recall-oriented task with two people collaborating under assumed roles). We did continue this work further by identifying more situations and defining other roles, but as I returned to UNC and continued working on my dissertation, I started moving in the direction of user-mediated collaboration. My dissertation provided a framework (among other things) for studying and supporting user-focused CIS. I have continued working on various aspects of CIS (both user and system sides) as a faculty at Rutgers University. In the meantime, I have also participated in a number of professional events around CIS, including half a dozen workshops – two of which I co-organized.

This book is a culmination of all of these experiences, and while they have made me biased on the topic, I have tried my best to incorporate others’ views as well. In the end, my hope is that those working in this domain, and the larger field of IR see this book as a record of modern day CIS research that has tried to incorporate many view-points and contributions to inform those looking for a comprehensive treatment of this topic, along with wonderful opportunities (and challenges) it presents.

Social search

Recently there has been a lot of talk about social search. Just last week I attended Microsoft Faculty Summit in Seattle, and one of the sessions was dedicated to social search. The panelists talked about social Q&A using Yahoo! Answers and Facebook status messages. It occurred to me during that session that all of the talks were really about social information seeking/retrieval, and not about social search. I raised this question after the panel presentations and Merrie Morris immediately agreed that everything that she was talking about social search was indeed technically social information seeking!

I kept thinking about what really is social search. Fortunately, the next day was the Social Media Day that allowed me to have more focused discussions with more specialized experts in the field. During the “birds of feather” lunch, I was at the social search table and through our discussions, it became clear to me that there are two ways of thinking about social search: a search done on social objects, or search done in a social network.

What is a social object? Two ways of defining it: (1) an information object that has social attributes such as name, age, gender, location, and (2) an information object created through a social construction, such as a Wikipedia article.

How is search done in a social network? It’s usually done by broadcasting information need to one’s social network. Think of people posting questions as their Facebook status updates.

Often social search is defined as a method of searching that takes into account connections among people in addition to connections among information objects. The above explanation/understanding of social search does hold with this definition.

Collaborative IR course at RuSSIR 2011

I had a fun time teaching a short course on Collaborative IR at the Russian Summer School of IR (RuSSIR) held at St Petersburg, Russia this year. I will share my experience in other posts, but for now, here’s a short description of the school.

The mission of the school is to teach students about modern problems and methods in Information Retrieval and Database Technology; to stimulate scientific research and collaboration in these fields; and to create environment for informal contacts between scientists, students and industry professionals.

The target audience of the school is advanced graduate and PhD students, post-doctoral researchers, academic and industrial researchers, and developers. RuSSIR/EDBT 2011 School will offer up to seven courses (in parallel sessions) and host approximately 150 participants. The working language of the school is English.

And here’s the description about the course that I taught:

The course will introduce the student to theories, methodologies, and tools that focus on IR in collaboration. The student will have an opportunity to learn about the social aspect of IR with a focus on collaborative IR situations, systems, and evaluation techniques.

Traditionally, IR is considered an individual pursuit, and not surprisingly, the majority of tools, techniques, and models developed for addressing information need, retrieval, and usage have focused on single users. The assumption of information seekers being independent and IR problem being individual has been challenged often in the recent past. This course will introduce such works to the students, with an emphasis on understanding models and systems that support collaborative search or browsing.

Specifically, the course will (1) outline the research and latest developments in the field of collaborative IR, (2) list the challenges for designing and evaluating collaborative IR systems, and (3) show how traditional single user IR models and systems could be mapped to those for collaborative IR. This will be achieved through introduction to appropriate literature, algorithms and interfaces that facilitate collaborative IR, and methodologies for studying and evaluating them. Thus, the course will offer a balance between theoretical and practical elements of collaborative IR.

The Coagmento difference

I often get asked how Coagmento is different than Google Docs, Diigo, etc. Actually there are a number of collaborative tools now available, and several of them have already achieved a good traction among end-users. So where do we fit in?

Well, from the beginning I have made it clear that Coagmento is not Google or anything close to it; it’s not designed to make it fast for “regular” searches. It’s also not merely an information exchange place like diigo or delicions, nor it is for simply creating a collaborative product like Google Docs. The real strength and the real difference that Coagmento brings is the ability to capture the process as it stores not only collective bookmarks, snippets, and final product, but also keeps track of various processes (search, share, interactions) that take place throughout the collaboration. This may not seem like a big deal, but it’s quite a powerful difference that Coagmento offers. This has significant implications for education, where it’s not the final product that we care about, but also the process that one went through to create it. This also goes along with a core objective of reference librarianship, where one tries to not only get an answer, but also shows how it was retrieved.

As we keep developing and testing Coagmento, we encounter more and more of such scenarios and applications. The reports of our lessons and findings will keep getting disseminated through several channels, including this blog.

Dissertation abstract

Following is the abstract from my dissertation titled “A Framework for Supporting User-Centric Collaborative Information Seeking” that I successfully defended on March 15, 2010. My committee included Gary Marchionini (chair), Deborah Barreau, Susan Dumais, Diane Kelly, and Barbara Wildemuth.

Collaboration is often required or encouraged for activities that are too complex or difficult to deal with for an individual. Many situations involving information seeking also call for people working together. Despite its natural appeal and situational necessity, collaboration in information seeking is an understudied domain. The nature of the available information and its role in our lives have changed significantly, but the methods and tools that are used to access and share that information in collaboration have remained largely unaltered. This dissertation is an attempt to develop a new framework for collaborative information seeking (CIS) with a focus on user-centric system designs. To develop this framework, existing practices for doing collaboration, along with motivations and methods, are studied. This initial investigation and a review of literature are followed by a series of carefully created design studies, helping us develop a prototype CIS system, Coagmento. This system is then used for a large scale laboratory experiment with a focus on studying the role and the impact of awareness in CIS projects. Through this study, it is shown that appropriate support for group awareness can help collaborators be more productive, engaged, and aware in collaboration without burdening them with additional load. Using the lessons derived from the literature as well as the set of studies presented in this dissertation, a novel framework for CIS is proposed. Such a framework could help us develop, study, and evaluate CIS systems with a more comprehensive understanding of various CIS processes, and the users of these systems.

Coagmento beta released

The first public version of Coagmento is finally out! It’s a pretty exciting and scary moment for me. I have been waiting for this for a long time (more than a year since I started working on Coagmento). I have done several iterations of Coagmento, many studies with different versions, and tried and tested it myself. So it’s great to be able to offer it to the world. At the same time, I know there are many things need to be worked out. People tend to compare everything with Google these days and that’s not a fair comparison; certainly not for systems like Coagmento. So naturally, I am anxious about the adoption.

To encourage people to try Coagmento and provide us feedback, we are even giving out prizes that include iPod Nanos and iTunes Gift Cards. More details can be found from CSpace (the space that you get once you login). I will try to keep posting more updates about the feedback and developments of Coagmento as the beta testing goes along. I believe there is much to be learned by all of us here – researchers, software developers, and educators alike. Stay tuned!