Information Technology Reference
In-Depth Information
Between 2005 and 2010 , knowledge - oriented search tools and exploration tools
will become widely available . Users major search tasks will probably switch from
data - oriented search to comprehension and interpretation tasks . Intelligent software
agents will begin to mature for citation context extraction and summarization .
Genomic maps will play more substantial roles in linking scientific data and
scientific literature . A synergy of data mining in genomic map data and scientific
literature will attract increasing interest .
Beyond 2010 , mapping scientific frontiers should reach a point where science
maps can start to make forecasts and simulations . Powerful simulations will allow
scientists to see the potential impact of a new technology . Further than that , we will
have to wait and see .
Many techniques have matured over the last 10 years, including automatically
summarizing multiple documents, automatic construction of ontology, and recom-
mending relevant references. Research has begun to touch the issues of predictive
analysis and how to deal with unanticipated situations. To what extent is scientific
advance predictable? What can we learn from the past so that we will be able to
better recognize early signs of something potentially significant?
I envisage the following two milestones ahead for mapping scientific frontiers.
First, recall the clarity of the conceptual structures demonstrated by Paul Thagard
as we have seen in Chap. 1 . Here are the requirements: at any point of time, the first
part of the input is the entire knowledge that has ever conceived by human beings,
the second part of the input is a newly proposed idea, the future system will be able
to let us know very quickly to what extent the new idea has been addressed in the
past and, if it is true, what areas of our knowledge will be affected. This process is in
essence what scientists go through so many times in their research. The key question
is how much of the retrieval, sense making, differentiation, and other analytic tasks
can be performed with considerably more external help.
Figure 9.12 illustrates how the publication of an article by Galea et al. in 2002
altered the holistic system of our knowledge on post-traumatic stress disorder. The
Galea article is six pages long. It cites 32 references. On the one hand, it requires a
substantial amount of domain knowledge to understand its validity and significance.
On the other hand, co-citation patterns indicate its special position in the landscape
of the domain knowledge. The diagrams show how the key contribution of their
work can be summarized at a conceptual level that sense making tasks can become
much easier and efficient.
The second milestone may be built on the first one to a great extent. The second
milestone is to externalize all the activities associated with scientific inquiries in
a form that can greatly integrate and inform scientists of their current situations
and paths that may lead to their goals. Figure 9.13 shows an illustrative schedule
of a fitness landscape of scientific inquiries. Each point of the landscape indicates
the fitness value of corresponding points on the base of the landscape. Many
scientific inquiries can be conceptualized as an exploration on such a landscape.
Some areas they find consistent information. Some other areas they may expect
to find contradictions. Some areas may be well defined, whereas other areas may
Search WWH ::




Custom Search