Hi everyone,

I had a person working on a school project sign up to AiDreams and he’s looking for people to spend ten minutes (that’s about how long it took me) testing his chat bot.

I wouldn’t ask this normally, but he is stuck in the manual registration system here and it’s probably going to happen too late.

He has until tomorrow, so this is my good deed for the day. If you would like to help him then please view this thread at AiDreams and help a budding chat bot maker in the making.

http://aidreams.co.uk/forum/index.php?topic=5943

Many thanks !

I suggested to Jake (the person Roger mentioned) over at the ALICE forums that he post here to ask directly. I also told him to mention me, and to ask you about mentioning him, to help allay any suspicions regarding his request. I think his project has some merit, so if anyone here is willing to take a little time to help a student succeed in his classes, I would be grateful, and I’m sure he would, as well.

Below are the results of very preliminary testing. Emphasis on preliminary. They are not intended to criticize at all.  All comments that follow are intended to be friendly and helpful, in a report form.  Nice job on the chatbot overall.

It would seem the chatbot is simply not finding its script: memory.chat, which is in the same directory.  That’s not much of a big deal.  This kind of stuff happens often with early testing of new chatbots.  Otherwise, the chatbot installed easily, and did not fail at any time, and it interacted very quickly. 

You: Why don’t you get out more?
Anonymous: I honestly don’t know.

This test was run under Linux with Open JDK Java 7.  The JAR file came right up as a desktop application, and worked well, with a textbox for input, a button for submit, and a textarea for a multi line log. 

I think this disconnect may not be a fault with the chatbot itself.  It is more likely, I think, related some configuration needed in the operating system, environment variable or whatever,  so that the chatbot could find its memory file.  The README.txt was followed carefully during installation.

Please note: This report is not final.  I will update it again, when I make some progress.  There are still a bunch of things left to try.  This is just a quick snapshot.

Report Update:

Progress has been made.  It is working perfectly.  The next thing I tried was running the JAR file from the Linux Terminal.  The desktop application came right up, but this time it obviously found its memory file. 

How to:
#: chatbotexperiment$ ./ChatbotExperiment.jar

Results:
You: What shall we discuss?
Anonymous: What’s your favorite movie?

You: How are you feeling?
Anonymous: I’m OK..

DEBUG REPORT:  The scrollbar disappeared once the log filled the entire page.  And the survey never showed up.  Is there a link to the survey?

Report Update:

Despite the minor scrollbar issue, the survey was finally reached at the end, and it was completed and submitted.  Minor issues are to be expected on artificial intelligence prototypes, simply because the priority is the A.I. over the interface.  However, the interface does serve an important purpose when testing, and should be fine tuned as needed.

This chatbot experiment deserves at least an A grade.