Dear Bruce,
Thank you very much. I want the system to
1. Perform a semantic search over the email corpus
2. understand the emails content
3. draw useful inferences from multiple emails(considering actual body
of the email + metadata like sender id ,reciever id, time,date
...etc).
I hope my first task can be done easily using PyKE. Now i am doing
the semantic search using LUCENE Java API and WordNet.
To build commonsence reasoning i have gone through OpenCyc. But it's
inference engine is not an open source.
Thank You very much.
KRISHNA
On 8/7/09, Bruce Frederiksen <
dang...@gmail.com> wrote:
> I have not done any work in forensic analysis of email messages, so it is
> difficult to recommend tools.
>
> So all I can do is provide some broad, general advice:
>
> - If you are new to Python, I would recommend that you learn it. You
> will find that it is generally much easier to write programs in Python
> than
> most other programming languages.
> - PyKE provides Prolog-like inferencing, but does not try to be a pure
> logic programing tool (like Prolog does). Rather, it is designed to
> allow
> you to use PyKE where logic programming makes sense, and to also use
> Python
> where standard imperitive programming makes sense; and to then be able to
> easily integrate these two very different programming methodologies.
> - I should think that PyKE would work well if you already know what
> patterns you want it to look for -- even if these patterns are quite
> complicated. It is not, however, designed as a self learning tool --
> meaning that it is not designed to be able to figure out the solution to
> problems on its own.
> - While PyKE should work well for email analysis based on looking for