The explosive growth in biomedical literature has made it difficult for researchers to keep up with the advancements, even in their own narrow specializations and to explore connections to their own work from other parts of the literature. LitLinker is a text mining system that incorporates knowledge based technologies, natural language processing techniques and data mining algorithms to mine the biomedical literature for new, potential causal links between biomedical terms. Click on the LitLinker icon to the right to try out LitLinker.

Text Mining Interface
LitLinker returns a complex set of data with connections between medical concepts that are new to users. Because these connections are new, one of the most important aspects of the LitLinker interface must be the ability to examine how the connections were generated. The interface must help the user understand the text-mining process and allow them to examine how the terms are connected in the scientific literature. In order for users to understand how the connections were generated an important aspect of the interface must be helping users understand the difference between the three types of terms and how they are each involved in the text-mining process. While helping the user form a conceptual model of how the connections were generated we must also keep the interface simple enough that it will not overwhelm the user. This is a challenge and a great opportunity to apply information visualization techniques to the text-mining process.

Metamap Evaluation
Metamap is a tool created by NLM that identifies medical terms in free form text. Although many system developers in biomedicine used this tool as part of their systems, few have studied how well it accomplishes the task of medical term identification in general. In this project, we conducted a study to evaluate the results of Metamap with medical concepts identified by a group of six medical researchers.

Information and the Cancer Experience

The goal of the information-and-the-cancer-experience project is to inform the design of new systems and services to support people receiving outpatient treatment for cancer. We begin with the premise that it takes work to be a patient in the current cancer care system. For example, people facing cancer must navigate the healthcare system, understand their evolving health status, and participate in their treatment. Many patients—regardless of socioeconomic status—encounter difficulties completing information-based tasks including managing their patchwork of treatment referrals, comparing different treatment options, remembering information from healthcare consultations, organizing information for future reference, locating information at the point of need (e.g., emergence of a side effect), or simply figuring out who to talk to about what. The prevalence of outpatient cancer care places additional burden on patients by leaving them to manage many tasks away from the treatment center. Our approach is to use socio-technical techniques to examine patient activities as work. A primary objective of our research is to inform the design of new support systems that reduce the cognitive load placed on cancer patients as they traverse the treatment system, manage personal health information and receive quality health care. Such systems can free up valuable time and energy that allow individuals facing cancer to focus more personal energy on the intangible tasks of health, healing, and wellness.

Results from a user study showed that breast cancer patients and their family members could find significantly more answers in a fixed amount of time, and were significantly more satisfied with their search experience when they used DynaCat than when they used either a cluster tool or a standard ranking tool. Subjects indicated that DynaCat provided an organization of search results that was more clear, easy to use, accurate, precise, and helpful than those of the other tools.