The Learning Chatbot

1
The Learning Chatbot

• Bonnie Chantarotwong
• IMS-256 Fall 2006

2
What is wrong with state of the art chatbots?

• They are repetitive
• They are predictable
• simple pattern matching set response
• They have no memory
• can lead to circular conversations
• They dont sound like real people

3
Proposed Solution

• Train chatbots on a corpora of conversations in
  order to mimic a given personality

4
Filtering the training corpus

• Need a lot of conversations containing query
  screen name
• Eliminate undesirable data
• phone numbers
• Addresses
• sensitive gossip
• Eliminate highly technical data
• since most tech problems are very specific,
  unless the bot was trained on a newsgroup,
  learned responses are not likely to be useful for
  tech support

5
Parsing the training corpus

• Extract messages from HTML
• Group together consecutive messages by the same
  screen name
• Simplify prompt messages
• !!!!!!!??????? -gt !?
• Ohhhhhhhhhhhh! -gt ohh!
• WhATz uP?? -gt whatz up?
• hahahahaha -gt haha
• Break prompts into word sequences (eliminating
  stop words)
• I took the cat to a vet -gt i, took, cat, to,
  vet

6
Constructing the CFD

• CFD conditions are prompt words
• FD samples are string responses, with numeric
  count indicating strength of correlation
• Example
• Cfdsleep.sorted_samples() -gt
• sleep is the best thing ever, are you
  tired?, maybe after I eat., hang on a sec.

7
Constructing the CFD
Simple Concept If a prompt is n words long,
then each word is 1/n likely to have caused the
response
i
1/3
Can we put mittens on it?
1/3
want
1/3
kitten
1/3
1/3
Me too, Im hungry.
food
1/3
1/6
they
1/6
What kind did they have?
1/6
had
1/6
good
1/6
1/6
at
restaurant
Original ConversationA I want a kittenB Can
we put mittens on it?A I want foodB Me too,
Im hungryA They had good food at the
restaurantB What kind did they have?
8
Using the CFD

• Problem
• Each word in a prompt is not equally likely to
  have caused the response
• More common words (such as I) are less
  indicative of meaning
• Solution
• Take into account the commonality of the word
  over all conversations
• Divide the weight of the word/response pair by
  the weight sum over all samples for that word
• Rare words are weighted more using a dynamic
  scale
• This improved quality of bot responses greatly!

9
Using the CFD - Example
CFD Cfdi (Can we put mittens on it?,
1/3), (Me too, Im hungry, 1/3) Sum
2/3 Cfdwant (Can we put mittens on it?,
1/3), (Me too, Im hungry, 1/3) Sum
2/3 Cfdkitten (Can we put mittens on
it?, 1/3) Sum 1/3 Cfdfood (Me too,
Im hungry, 1/3), (What kind of food did they
have?, 1/6) Sum 1/2 Cfdthey (What
kind of food did they have?, 1/6) Sum
1/6 Cfdhad (What kind of food did they
have?, 1/6) Sum 1/6 Cfdgood (What
kind of food did they have?, 1/6) Sum
1/6 Cfdat (What kind of food did they
have?, 1/6) Sum 1/6 Cfdrestaurant
(What kind of food did they have?, 1/6) Sum
1/6 Responses (how is your kitten?)
(Can we put mittens on it?, (1/3 / 1/3)
1) Responses (the food was good) (Me too,
Im hungry, (1/3 / 1/2) 2/3), (What
kind of food did they have?, (1/6 / 1/2 1/6 /
1/6) 4/3
10
Using the CFD

• Responses (the food was good) (Me too, Im
  hungry, (1/3 / 1/2) 2/3),
• (What kind of food did they have?,
  (1/6 / 1/2 1/6 / 1/6) 4/3
• Given the CFD, the response to any prompts
  containing food and good will give back What
  kind of food did they have?
• Problem This can lead to redundancy
• A The food was good
• B What kind of food did they have?
• A Didnt you think the food was good?
• B What kind of food did they have?
• Solution Store an FD of used responses, and
  dont use them again
• A The food was good
• BWhat kind of food did they have?
• A Didnt you think the food was good?
• B Me too Im hungry

11
What if we have no responses?

• Because
• Weve never encountered any of the prompt words
• Weve used up all relevant responses
• We can
• Find a random response
• Enhance randomness by favoring unlikely responses
  (near the end of association lists) to reduce
  redundancy
• Fabricate a response based on pattern matching
• Select a response from a default list of
  responses (i.e. Lets talk about something else,
  I dont know anything about that
• All my bots implement 1 2, and one bot
  (bonnie) also implements 3 4

12
Interactive Webpages are not trivial

• Especially if you want to retain some memory of
  the past
• First CGI problem A new bot is created with
  every web prompt - all memory is lost
• Solution
• Write all bot state changes to a file, including
  used responses.
• Run this file with every prompt, and reset it
  when a new conversation starts
• The bot loads the huge CFD all state changes
  from scratch with EVERY call. Slow, but works.

13
Interactive Webpages are not trivial

• Memory Self-modifying code

14
Interactive Webpages are not trivial

• Memory Self-modifying code

15
Demo
http//ischool.berkeley.edu/bonniejc/