I've been hunting around to see if I can find a working implementation of PARRY, but unfortunately nothing seems to be available (that is working at least). However, I did locate some PARRY source code at the AI archive at CMU...
http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/classics/0.html You can see what might be the original PARRY code here, as well as a few different versions of Eliza. As we've discussed Eliza has been implemented many different times, largely because it is a relatively simple method that is very nicely described in the original Eliza paper. PARRY, on the other hand, is a much different sort of chat-bot in that it includes a lot of very detailed world knowledge in order to (in effect) create a cognitive model of a paranoid schizophrenic. So, that isn't so simple to encode in a program, and it can't really be completely described in a paper, so I guess that's why we don't have any working versions of PARRY right now (or at least one that I could find). In any case, at least we can see the source code, and who knows, maybe it would be possible to try and re-implement PARRY from the code we have available? And of course, if anyone runs across a working version of PARRY, please do let us know! Enjoy, Ted On Mon, Nov 9, 2009 at 10:40 AM, Ted Pedersen <[email protected]> wrote: > Greetings all, > > Our next NLP group meeting will be on Tue Nov 10 at 2pm (in HH 321, > the CS conference room). We'll be discussing the Parry paper (by > Colby) in more detail, and also get started on talking about the > Turing test, by focusing on Turing's original paper. Both of these > papers are available in the Files section of the Yahoo group. > > Our next meeting after that will be on Tue Nov 24, and will feature > your Eliza programs! So, you should be finishing your alpha version > (with 10 keywords and 3 transformations, nicely explained to the user > via perldoc) in the next few days if at all possible, and then your > beta version (with 20 keywords and 10 transformations, nicely > explained via perloc) by this Friday Nov 13. I'll run those and give > some advice on what to do next, and then you should have your final > Eliza program ready by Friday Nov 20. I will again review those and > then share them with everyone in the group so that we can each run > everyone else's program, and we'll discuss what we observed in our Tue > Nov 24 meeting. I will ask you to send me at least 2 interesting > sessions you had with each Eliza program by Monday Nov 23, and then > I'll print those out and we can use them to discuss the programs in > our meeting. I will also print out your source code and perldoc so > that we can discuss that content as well. > > As to writing, once your Eliza summary has been finished and approved > by me, please start to work on a summary of the Parry paper (the Colby > one) and then after that you'll do a summary of the Turing paper. We > should have all of those finished before the Tue Nov 24 meeting, since > that will be our wrapup discusion of Eliza, Parry and the Turing Test > (and we'll look at each other's summaries during that meeting too.) > > BTW, if you haven't voted in our Turing Test poll, please do that asap! > > See you this Tuesday - please do let me know of any questions! > > Thanks, > Ted > > -- > Ted Pedersen > http://www.d.umn.edu/~tpederse > -- Ted Pedersen http://www.d.umn.edu/~tpederse
