Attribute / Relation – Personal / Social January 12, 2007Posted by Andre Vellino in Digital library, Information retrieval, Social networks.
I am taking my queue from the first entry in Peter Turney’s new blog regarding relations and attributes. In my neck of the woods (information retrieval in digital libraries) the corresponding dilemma is “the individual” vs. “the collective”. I think that people are like words in this regard: who they are (or “what they mean”) depends essentially on their (m-to-n) relationship to one another.
I’d like to investigate a web portal environment that enhances the search experience of scientists who use the CISTI digital library. This portal needs to catering to physicists, chemists, doctors, environmentalists, who want to browse or retrieve journal articles. But not in the way that physical libraries are currently used (catalog searching and database searching, primarily.) I have my doubts about whether Google Scholar is the right model either. My guess is that Citeseer has made a pretty good first stab at what a scientific e-library should look like. Except that I think it’s important to know the identity of your user not in attributional terms but relationally.
The scientific categories (nuclear physicist, environmental bio-chemist, etc.) that apply (as attributes) to the researchers who will use this portal belong is only one superficial facet of their identity. Other facets that matter are: what organization do you belong to (private sector / academic / government) and within that, what institutions are you affiliated with, what is the reputation of that intitution, and how authoritative is the information that has it imprimature. When added to citation information about the articles, the authors and the journals, not to mention the content of the articles, I think we’ll get a much better picture of how we can help the reader find what she wants.
These things matter because they speak to the reliability of the information – how much can it be trusted. The “interests” of a scholarly writer are not just about what the writer cares about but also about where his or her allegiances are. For example, if I care about the impact of CO2 emissions on the climate and I want to search an on-line library for important articles on the subject, it probably matters to me whether you (the author) are employed by Exxon or by CalTech (although even that information might be hiding the truth, namely that even though your research is at CalTech it is actually funded by Exxon and your work might or should be suspect.)