Arrowsmith Home Page

Informatics Tools for Scientific Discovery and Collaboration

University of Illinois at Chicago
UIC Psychiatric Institute
1601 W. Taylor Street (corner of Ashland and Taylor)
Chicago, IL 60612
September 3-5, 2003
To register or further information: contact , 312-413-4581

Wednesday, Sept. 3, 2003

2:00-6:00 pm. Tutorial, Practice and Question-Answer Session. For people who have not had previous tutorials: basics of conducting, evaluating and recording Arrowsmith searches. For those who are comfortable with the basics: Time for questions and running (or re-running saved) searches by the field testers under the eyes of the instructors. The Wednesday session is intended for Arrowsmith Project personnel; others may attend if space permits.

6:00 Dinner.

Thursday, Sept. 4

9:00 Welcome. Information is Money. Earn it, save it, invest it; and retire rich! Neil Smalheiser. Powerpoint slides

9:00 Basics of searching the biomedical literature. Medline fields, Medical Subject Headings, controlled vocabulary, UMLS, Metamap, semantic categories. Basics of PubMed. How it converts a user-typed query to the “real” query. Options for searches. Clinical queries. Augmenting PubMed: link-outs to the web (online papers, author/affiliation home pages, and links to Google searches); visualizing, clustering, filtering and mining search outputs. Searching other literatures besides Medline (Psychinfo, ERIC, EMBASE, Firstsearch, etc.). Alerting services: Pubcrawler. Other online search tools especially for biomedical information. Web of Science for citations. E-Prints and full access to the literature. Neil Smalheiser.Powerpoint slides

10:00 Strategies of searching the biomedical literature. Recall and precision. Techniques for broadening and narrowing searches. Relevance feedback, interactive search strategies. Don Swanson.Powerpoint slides

10:30 Coffee Break.

10:45 Introduction to complementary literatures and Arrowsmith. The ABC model of complementary literatures. Types of complementarity. Do the sizes of sets A and C matter? Rationale for filtering the B-list through a large stoplist. Highlighting ABC terms. Hypothesis testing and generation. Subject headings as a filter and means for B-term ranking. How search strategies influence Arrowsmith searches. Don Swanson.Powerpoint slides

11:30 The UIC Arrowsmith site. Different modes of (or reasons for) discovering information across two or more literatures. Examples. Existing B-term filters, planned filters and alternative output modes. Link by title words and phrases, vs. terms in abstract or full-text. Alternative ways of linking literatures: UMLS concepts, MeSH headings, affiliations. The open-ended (one node) search. Linking Medline to other bibliographic databases. Can one automate Arrowsmith searches, or automatically detect promising inferences? An alerting service to flag new papers that bridge disciplines. Neil Smalheiser.Powerpoint slides

12:30 Lunch.

2:00 Structure, scope and aims of the Arrowsmith Project. Neil Smalheiser.

2:30 Author linkages. Author disambiguation and linking literatures by author to study author publication behavior and collaborative networks. Monotone Boolean functions in information retrieval. Automatic recognition of types of web pages and extraction of information into database form. Vetle Torvik.Powerpoint slides

3:30 Extensions of the Arrowsmith approach to bio-informatics: gene-gene associations, finding sequence overlap, using microarray data, images or other supplemental data to link literatures. Making inferences across linked research (basic and clinical) databases. Vetle Torvik. Powerpoint slides

4:00 Coffee Break.

4:15 More than information retrieval: The active role of the Arrowsmith grant in following-up searches with new experiments and new collaborations. Open source lab notebooks. Software to enhance collaborative thinking and creativity. Online knowledge environments. How can informatics lead to new models of experimental design and scientific collaboration? Neil Smalheiser. Powerpoint slides

5:00 Information-gathering and collaborative behavior of interdisciplinary scientists. Carole Palmer. Powerpoint slides

6:00 Dinner.

Friday, Sept. 5

9:00 Feedback from field testers. Introduce themselves and their lab projects. Give and reconstruct examples of searches, experiences and suggestions. Open mic.

10:30 Coffee Break.

10:45 Interface design. How should a website be designed to optimize usability and a sense of community? The evolving Arrowsmith interface. Ruth West.

11:30 How to record and evaluate Arrowsmith searches. Vetle Torvik. Powerpoint slides

11:50 Arrowsmith information diary: Linking your information activities to your research process. Melissa Cragin, Tim Hogan. Powerpoint slides

12:15 Lunch.

2:00 Introduction to databases. Medline as a flat-file vs. Medline as a database suitable for querying. Why the WWW and even your own experiments can (and should) be thought of as a database. Databases for specialized data formats. What every neuroscientist should know about database design and data mining. Data and metadata. Ontologies. Visualizing massive datasets. Vetle Torvik. Powerpoint slides

3:00 Introduction to computing. Open source software tools. Wrappers. Introduction to supercomputers and parallel computers. Processing chips, graphic chips. Orders of magnitude in computation. The Grid. Amanda Grethe. pdf file of Powerpoint slides

4:00 Coffee Break.

4:15 Basics of Information Retrieval, Natural Language Processing and Information Extraction. Overview of research in medical informatics to automatically extract biological information from full text of scientific papers. Structure of the Web. The semantic Web, the deep Web. How does Google work? Search engines, metasearch engines, query processing. Clement Yu. Powerpoint slides

5:45 Questions and general discussion.

Last modified: Tue Oct 28 13:29:17 CST 2003