The Association for Library and Information Science Education (ALISE) and Online Computer Literacy Center (OCLC) have awarded $15,000 to Syracuse University School of Information Studies (iSchool) Assistant Professor Bei Yu to fund research in information seeking in community-based question-answering services.
“Given the new directions in which information seeking has been moving at an increasing rate, I believe that the proposed research into information-seeking behaviors in virtual reference in digital libraries will provide answers to many emerging questions if information seeking is going to flourish in the social web environment,” wrote iSchool Dean Elizabeth Liddy in support of the grant.
The research will primarily address how people seek information in computer-mediated communication (CMC) systems, whether or not the process has changed from the traditional face-to-face environment and what improvements can be made within the environments. These services include popular sites like Yahoo! Answers and WikiAnswers as well as question-answering services conducted using text-messaging on mobile phones.
“Digital reference services hold the promise of allowing library users to submit reference questions to a professional librarian when and where those questions arise,” wrote Howard Turtle, director of the Center for Natural Language Processing at Syracuse University. “The interaction necessary to support these services is much different than that of a face-to-face reference interview. If we are to build effective tools to support digital reference we need to understand the strengths and weakness of this style of interaction and the proposed research will contribute directly to this understanding.”
The first part of the research will be conducted with discourse analysis, wherein a data set in the form of a dialogue between an information seeker and information provider is annotated and then analyzed in order to determine which information-seeking processes are effective or ineffective during virtual reference interviews. This will contribute to building a predictive model showing which processes lead to successful information dialogue. The end result is to make recommendations on building a useful interactive web interface and discover a process by which to use social media as an information resource.
The second aspect of the research will use machine-learning experiments, wherein annotated data is used to train text classification systems in order to automatically identify the discourse. This will provide the proof of concept of the above dialogue model, by confirming that there are certain identifiable trends within this dialogue that can be reflected by a computer algorithm. The outcome is to create a new measurement for evaluative virtual reference services, new data attributes for information extraction and retrieval algorithms, and a prototypical dialogue model for constructing fully-automated dialogue systems.
Though Yu is listed as the primary investigator, the research proposal is based on the dissertation topic of iSchool doctoral candidate Keisuke Inoue. The award will primarily fund graduate assistants as well as Inoue during the course of the research.
Before joining Syracuse, Yu was a postdoctoral researcher at the Kellogg School of Management at Northwestern University. She earned a Ph.D. from the Graduate School of Library and Information Science at University of Illinois at Urbana-Champaign and holds both BS and MS degrees in computer science.