ArsRSS is a news feed service, intested in
collecting data about how the service is used. It
is an ongoing project, adding to the types of data it collects as well as to the
mechanisms for collecting and analyzing the data. The service is based
on the parsing of up to 120 news feeds every four hours between 8 a.m.
and midnight Eastern Time. This is its core and it keeps track
of every word in the feeds, how often they occur, and in which feeds.
ArsRSS is interested in the frequencies with which terms are queried, and
in which news feeds supply responses to queries and how often. It is interested in where
the queries come from and how often, in the relationships between what people ask and
what the news feeds provide.
In addition to its query service, ArsRSS has a news feed Reader which enables its users
to accesss from its cache the 100 to 120 news feeds which form the basis of
its activities. Which feeds are read by visitors and how often is also of interest
to ArsRSS.
Finally, as the project has developed, an underlying concern has become clear, namely
what it is that we can learn about users from just the basic tools and infrastructure
of a web site. This gets us into questions of the uses of web architecture for profiling.
These panels--
basics,
queries,
feeds, and
words--provide
some information about the kinds of data that
ArsRSS is collecting.
places is a pop-up window, using google maps, that shows
where queries come from and in what numbers. As the project continues, and as more data
becomes available,
ArsRSS will develop additional browser-based mechanisms to
make the data available.
Opportunities
