Transcript Service Learning

Marvin - A Conversational
Agent Based Interface for
the Study of Information
Sciences
Nives Mikelic Preradovic, Sanja Kisicek, Damir Boras
University of Zagreb, Faculty of Humanities and
Social Sciences, Department of Information Sciences
[email protected], [email protected], [email protected]
Outline
 Who is Marvin?
 Marvin’s knowledge base and reply
structure
 Evaluation
 Future work
Who is Marvin?
Who is Marvin?
 Has a "brain the size of a planet"
which he is seldom able to use
 No task he could be given would
occupy even the tiniest fraction of his
vast intellect
Why Marvin?
 Aim was to build a virtual character that:
 Is capable of engaging our undergraduate
students in a meaningful conversation
 Has the ability to parse NLQ and, by referring to
a knowledge base, generate NLA
 Will increase the students' interest in the Study
of Information Sciences, ease the search for
particular information and boost satisfaction with
the presentation of information
 Aim was not to
 design a tutor that would teach particular
subject matter
Marvin’s knowledge base
 AIML: category (pattern,template, context:
that, topic)
 Reqursion:
 Marvin produces the same reply to different
query formulations that share the same or
similar meaning (teach-lecture-instruct)
 AIML templates activated by the
appearance of a keyword anywhere in the
input
 Keywords: the most probable user's queries
 Knowledge base
Marvin’s reply
 Tries to boost communication with the
user posing the questions
 Topic change triggers the content of
the reply
 Marvin attempts to follow the normal
flow and nature of the communication
 Marvin’s replies are enhanced with
comments designed to convince the
user that Marvin is a student himself
Evaluation
 The database was checked for accuracy,
interpretation and relevance to the types
and the level of questions being asked
 50 inputs and answers per evaluator / each
evaluator spent 30 minutes on average
chatting with Marvin
 Two perspectives: competent information
scientist / pretend bachelor student
perspective
 Limitations: multi clause units, exchanges
that range over more than one turn
Evaluation
 the depth of knowledge too narrow to
cope with open-ended conversations
with humans
 database should be improved with
more factual world knowledge
 Currently: no actual knowledge of what
it is talking about, avoidance strategy
 He or it?
Future work
 Modify Marvin to take on the
personality and knowledge base of
the key information theorists
 Sensitive to the knowledge levels of
student users and opened to
explanatory questions
 Use Wikipedia information to build its
conversations and offer links to
articles in the field of IS