Hi all.
I'm an undergraduate in the UK, in the final year of my course. For my
final year project, I am attempting (at the suggestion of Ian Clarke) to
add a search capability to freenet.
The idea was best described by Ian himself - his email is included
below.
Right now, I am at the stage where I need to get hold of the source and
get to grips with the relevant bits. What version of the source should I
be looking at?
Is there any documentation of the overall structure of the code? Are
there any significant pieces of infrastructure in place that might
seriously affect this project that I might not know about?
What are peoples views on the idea generally? does it sound feasible?
Thanks, ned.
Ian's email:
Certainly one interesting area might be investigating how the current
lexographic measurement of key-closeness could be replaced by a
comparison which allowed metadata (pairs of keys and values) to be used
as a key in Freenet, and a heuristic search to be conducted for that key
with a SQL-style search term such as:
("name" = "Fred") AND (("age" > 33) OR ("age" < 10))
Fuzzy logic would be used to determine how closely a given key matched
this criteria (possibly using "edit distance" to measure how closely two
strings match), and a request routed on that basis in the normal freenet
style.
Rather than search for a specific piece of data, the request would
collect a "top-ten closest matches" as it passed along the search path
which could then be returned to the requestor, who could select which
data she wants and request that using the conventional Freenet mechanism
for requesting data when you know the exact key.
_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://lists.freenetproject.org/mailman/listinfo/devl
