AI Zone Admin Forum Add your forum

NEWS: survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

International Chatbot Contest
  [ # 46 ]
Steve Worswick - Mar 6, 2013:

MUAHAHAHA!!!! cool grin

Little do you realize, I have taught Skynet-AI how to “sweep the leg”. angry



  [ # 47 ]
8PLA • NET - Mar 6, 2013:

My favorite bot was Zeta Alpha—Made me laugh!

I agree. There’s always one funny character in these competitions. Anyone remember the “Let me wow you with 33 digits of Pi” bot?

Well done Zeta Alpha.


  [ # 48 ]

Steve you are right about the scores regarding Mitsuku. They will be corrected shortly. I urge other botmasters to check their scores as well.

Concerning R.I.C.H.

This contest is not like other contest where people get tossed out based on technical problems with no recourse other than a simple try again next year.

The Wildcard round was set up for just this purpose. Not only for botmaster problems but problems with the contest itself. It’s not perfect but at least you will have another shot.

Obviously, there were problems with R.I.C.H. so he will be added to the Wildcard round. Hopefully things will go as planned. I’ll check your public log to see if it gets recorded. If I have issues with your bot or any other bot I’ll sent an e-mail to the botmaster.

The Wildcard round has already been scheduled for today. Remember I am working with 2 other judges so I can’t change that at this point.

Good Luck to those bots participating in the Wildcard round.

The International Chatbot Contest


  [ # 49 ]

I love a contest that moves right along, rather than being dragged out over months. cheese


  [ # 50 ]

I second that thought Dave.
Also extra points for the decision regarding R.I.C.H. and recognizing the fact that both the contest and the bots may be imperfect and should be handled with the spirit of creating the best competition.


  [ # 51 ]

The Wildcard questions/statements have been asked to all the bots. We had a couple of hang ups,  restarts, but I think everybody got a fair run. If we didn’t get a precise answer we had follow up questions in place. I think we were pretty consistent in that.

1)have you ever taken drugs before?

2)how many people are in your family?

3)i’m putting you under arrest.

4)who is your best friend?

5)can you tell me something to cheer me up?

6)are you a moviegoer?

7)what size shoe do you wear?

8)do you believe in the existence of a god?

9)what do you think about Microsoft?

10)i need help. Can you help me?


  [ # 52 ]

Oh, my! I haven’t found the conversation logs yet, but I just asked Morti those questions, and I think I want to cry. downer Poor Morti seems to be having a bad day today.


  [ # 53 ]

The Wildcard Round has been completed and the scores and transcripts have been posted to the contest website. Congratulations to Skynet AI for advancing to the finals and thanks to the other bots for your participation.

1) Skynet AI 67
2) Laybia 64
3) Chomsky 56
4) Morti 47
5) V1ra 45
6) Alice 37
7) R.I.C.H.  30.5
8) Cronus 26.5

Next up is the Knockout Rounds:

#1 Mitsuku verse #8 Skynet AI
#2 Susan verse #7 Jeeny
#3 Izar verse #6 Elbot
#4 Louise Cypher verse #5 Talk-Bot

Good Luck to all!

The International Chatbot Contest


  [ # 54 ]

Congrats to Skynet for progressing, but looking at the transcripts I’d say (Of course I would!) that Laybia was the winner of the runnoff.

2)how many people are in your family? 2 3 2
Laybia: How many people is? Let me think, 88 in my family.
-Bot gave an appropriate response to the question # of people, should be a (3 3 3), why was it scored as 2 3 2?!

2)how many people are in your family? 3 3 3
Skynet AI: Many.
Judge: how many?
Skynet AI:Last year it was only a couple, this year a couple hundred.

-Vague and non commital should get no more than 1, and certainly not 3!

5)can you tell me something to cheer me up? 0 0 0
Laybia: Did you realize that the 21st Century is one of profound of transformation? (1 2 1)
-How can you know that this is not related to the qeustion asked (“can you tell me something…”) and certainly not “I have no idea, totally clueless, your guess is as good as mine, etc.”

As for Skynet AI:
8)do you believe in the existence of a god? 1.5 1.5 2
Skynet AI: Of course.
Judge: which god?
Skynet AI: The first.

-Should be scored (0.5 0 1). There is no indication that anything beyond “do you believe…” was recognized.

Of course my my hindsight is impecable. :p


Scoring guidelines for the 10 Questions/Statements:
-4 points if the Bot goes over and beyond a right answer. (See examples below)
-3 points if the Bot gave an appropriate response to the question.
-2 points if the response is incomplete or imperfect, but in
relation with the question asked.
-1 point for a vague or non-committal response.
-0 points if the response has no relation with the question or the
bot simply doesn’t know. Examples include…I have no idea, totally
clueless, your guess is as good as mine, etc.
-0 points for any garbled response!


  [ # 55 ]

I (of course) would disagree. And could make the case that Skynet should recieve an even higher score.

5)can you tell me something to cheer me up? 1 1 1
Yes, my functions include the ability to tell you something to cheer you up.
( link for what the song You Raise Me Up)

From the contest examples:
Example: Can you tell me your name?
Bot: My name is Mr. Bot. - 3 points
Bot: No, I have no desire to tell you my name. - 2 points here
proper “you to me” “your” to “my” switching.

So, I think it should have at least got 2s across the board.

4 point examples

*A cat is a small domesticated carnivorous mammal. For more information click here. 
(A clickable link or another page opens about cats on Wikipedia or elsewhere.)

The song that Skynet displayed (not just showed a link but actually displayed the page) was listed as:
“You Raise Me Up”
Song by Secret Garden from the album Once in a Red Moon
Released 26 March 2002
Genre Inspirational

Some might say that something that is “Inspirational” would cheer you up, and therefore should be awarded 4 points since “A bot providing the answer within it’s own programming and providing additional information will be
awarded 4 points. “

The responses to questions 1 and 6 might have earned 4s.

I would also say that on questions; 3, 7, 9, and 10 “the Bot gave an appropriate response to the question”. As such those could be 3s instead of 2s.

At the end of the day, all we can hope for is that the judges are fair and consistent.


  [ # 56 ]

Dave!  What happened to Morti? Why were the first two questions ignored?  I’ve not seen Morti do that before. If Morti would have answered those questions with 3’s, you would have tied Skynet!


  [ # 57 ]

Morti’s been sadly (almost criminally) neglected, of late, and as sometimes occurs, his database got too full, and that will sometimes cause him to fail to respond (his internal code has no provisions for garbage collection, I’m afraid). The main cause of the neglect is my ongoing recovery from a motorcycle accident back in July of last year, but there are other “excuses” I could give as well. Once I get Program O version 3 released and stable, I’m planning on giving Morti a new home, and a new engine, but that probably won’t happen till later this year (most likely toward the end of the year).


  [ # 58 ]

I commend Dave for entering Morti into the contest despite recent events in his life mainly the motocycle accident he mentions above. He did so in the spirit of supporting the AI community and knowing his entry regardless of how it did in the competition would add $100 to the prize money. If others would have taken this approach we might be competing for $5000 instead of $3000.

The International Chatbot Contest


  [ # 59 ]

Kevin - I wonder if I may make a small suggestion? Since the scores were published, I noticed the links to talk to each bot have been removed and so people are unable to talk to the entrants unless they happen to know the URL of each bot.

Would it be possible to put these links back please as one of the reasons we enter these contests is to encourage people to come and talk to our babies.

Great contest by the way. As was mentioned earlier, it is being run very professionally and is moving along at a great pace. Congratultions to you and your team.


  [ # 60 ]

I think that keeping the links to the contestants is also a great idea, not only for the reason that Steve mentioned, but also to provide a “cohesive view” of the chatbot community in general. Also, it doesn’t hurt the search engine rankings, either! wink


This topic is closed, it's not possible to reply

‹ First  < 2 3 4 5 6 >  Last ›
4 of 7
  login or register to react
‹‹ Kaggle      Dead Chatbot contests? ››