NeoRef (Scholarly Communications, Annotations,
NeoRef is an archive for any digital material including articles, research notes, books, genetic sequences, concepts, annotations, reviews, and claims. Areas of emphasis include development of global paradigms for colloboration and sharing, annotation models, open publishing models, optimization of the author self contribution process (user interface, selection and assignment of index terms from controlled vocabularies), automatic metadata extraction from existing materials, optimization of user interfaces for searching from very large content sets (federated universal open archives), and extending metadata to include "claims" to capture higher level knowledge representations about linkages between content items (in particular extending the ScholOnto work to the domain of bioinformatics). A recent writeup about vision for the global colloborative sharing paradigm (oriented for annotating scholarly work) is found here. An example implementation is documented in our 2008 YouTube video on NeoNote. We also study the technical and policy issues involved with freely available archives of all scholarly literature that allow for the storage, searching and retrieval of information without the need for the traditional framework of publishers and review systems. A general NeoRef presentation is NeoRef Scholarly Communications
We are studying information seeking, primarily in academic settings. We have completed a national survey of information seeking at five universities. We are conducting interviews with individual faculty and laboratories about their workflows and practices. We are also interested in information seeking interfaces, and are working with UNC and TRLN to design better library catalog interfaces. We have obsever studies and log analysis studies examining how people use university library catalogs. A particular area of interest is combining faceted based searching with text searching, and also dynamic topic clustering.
Visualization and User Interface Design
Development of computer workstations as the display interfaces for the practice of medicine, and in particular, radiology. Previous work includes design principles for radiology workstation design, and specific designs and implementations (chest CT, mammography, 3D). Current interests include (1) investigating what choices of mental models and roam and zoom techniques are most effective when using images larger than the display resolution, examining what visualization and interactive features are effective (or not effective) for mobile devices like iPhone and Android, and (3) effective CHI techniques for dynamic searching of very large sets of content items (ie. universe of scholarly publications, or at least library catalogs). Some examples tools our lab has built include ICIS--Interactive Comments in Schema, JAR--Journal Author Rights, PHR--Personal Health Records.
Ultrastructure: Working with Morgan Giddings and Jeff Long to study the application of Ultrastructure, a novel notational system to represent knowledge in a flexible way using relational databases, to bioinformaticss to represent and integrate current and future information information from all genome sciences domains. We have also working on development and design of representation structures (primarily databases) for a number of research projects at UNC including mouse databases, proteomics databases, plant comparative genetics, schizophrenia. An example tool is TAMAL
Dr. Hemminger helped found The Center for Research and Development of Digital Libraries (CRADLE) in 2002, and currently serves as the organizer of the group. This project brings together many active digital library researchers and projects at UNC and worldwide. He is also interested in Electronic Theses and Disserations. He has developed a ETD for SILS scholarly literature (part of NeoRef), and chairs the campus group that developed ETDs for UNC-CH, and continues to expand services provided by the library for supporting ETDs on campus.
I have proposed a methodology for digitally capturing both the content and entire exhibits of museums, so that they can be visited in a virtual reality setting. The virtual reality representation is unique in that the environments and objects are captured and displayed in way that is both visually compelling (photographic quality), and spatial accurate (millimenter measurements are possible). This work is in conjunction with 3rdTech, who developed the DeltaSphere digitizer. Virseum
Digital Imaging and Communications (DICOM) standards work. Dr. Hemminger chairs Working Group 11 (Display), and is an active member and author in Working Groups 15 (Digital Xray, Mammo, Computer Aided Detection), Working Group 17 (3D), AAPM Task group 18 (Electronic Display). For a listing of existing DICOM international standards and new standards under development see the DICOM website at NEMA.
Integating disparate databases. Handling very large databases. Primary emphasis is on biological databases (genetics, protoemics, metabolomics).
Knowledge discovery via data mining of large and integrated databases. Application of statistical pattern recognition and feature analysis techniques to the analysis of genomic and proteomic databases.