Project:

Topic/trend detection systems composed of crawler and extractor, topic
extractor and topic detector. My project will focus on the crawler and
extractor, which will restrict the web data from media resources. I
observed that the existing approaches including the focused crawler
intrinsically include Slow Intelligent System (SIS) modeling principles
including Enumerate, Adapt, Eliminate and Concentrate. Time controller
model has not been introduced yet to the existing approaches. For the
first phase of my project, I will select the most proper solutions from
focused crawlers, other adaptive crawlers and extractors with respect to
closeness of SIS models. The selected approaches can be modeled as
relevant SIS patterns (e.g., Classifier from focused crawlers can be
modeled Enumerator or Adaptor) and attributes for the extracted models.
Then finally the informally modeled patterns can be represented as
PetriNet model or relevant modeling, which can be exported as an XML
format.

I will:
 1) Investigate patterns for the existing approaches to restrict the web data
 2) Model the patterns of 1) in SIS principles by utilizing PetriNet or
 other relevant modeling tools.


Comments:
What patterns? You need to be more specific by
giving perhaps one example of a patter you found.