Hi everybody,

We are interested in investigating methods of distinguishing causal and
acausal relations based on time (rather than the more traditional method
of conditional idependence).

For this reason I am looking for a rather peculiar kind of data: I am
looking for causal relations in temporal data. By temporal I mean the data
should have produced at regular intervals (every second, every day, etc.).
For a causal relation to exist, the data should come from a single
source, and should have the value of at least a few related variables. (A
time-series dataset is not very useful because a single variable can
hardly be used for causality purposesd).

The above two rquierements mean that "census-type" data whihc abound on
the Internet are not suitable. First because there is no temporal order
among each field, and second because they come from different "systems"
(peopel, things, etc).

I wonder if anybody on this list knows of such temporal data, or of any
application program that can process such data.

I would be grateful for any help.

Kamran Karimi
karimi@...
http://www.cs.uregina.ca/~karimi