Information Technology Reference
In-Depth Information
and producing speech, gestures, and facial expressions. Nevertheless, they often fail to
converse in great depth and hence to mutually establish a topical talk with their human
opponent. In addition, many ECAs lack in simulating a sense for the adequacy of certain
topics during dialog. To remedy these weaknesses, the artificial interlocutor needs to be
aware of ongoing and potential conversational topics like humans.
To provide conversational agents with artificial, humanlike topic awareness in ev-
eryday interactions two main tasks need to be automatized: First, the detection of
topics raised in ongoing natural language dialogs and second, the adequate integra-
tion of the resulting topic information into the agent's underlying system architec-
ture. This paper introduces an approach tackling both tasks: We show how to connect
well-established linguistic information retrieval methods with benefits originated from
collaborative work provided by Wikipedia to automatically detect dialog topics. Addi-
tionally, we present how to utilize the obtained information to improve the conversa-
tional abilities of virtual computer characters regarding topic handling.
The rest of the paper is organized as follows. In the next section we introduce our
notion of dialog topics establishing the basis for the present work. Subsequently, the
several processes of our automatic topic detection approach are described in Section 3.
Thereby we especially emphasize the application of collaborative knowledge provided
by Wikipedia. Section 4 highlights the embedding of the resulting topic information into
the existing architecture of the conversational agent Max. As a result, we contribute in
emulating humanlike topic awareness in artificial agents as described by means of our
dialog scenario in Section 5. Moreover, we present how to evaluate our model in the
near future. In Section 6 we give an overview of related work before closing the paper
with a short conclusion and discussion.
2
Introducing Dialog Topics
Assuming dialogs to be face-to-face conversations between two partners, a dialog topic
emerges from a
joint activity
performed by both interlocutors [6]. That is, considering
single utterances to specify a dialog topic is insufficient as they do not have topics in
isolation. They rather provide topic suggestions [7]. However, the topic formulation of
the particular topic is done at different levels of abstraction and from different subjective
positions [2]. Speaker A, for example, might categorize a dialog about Whiskey and
Brandy using the term “Alcohol”, whereas speaker B might choose the term “Drinks”
or “Spirits” referencing the same topic. According to this, we define a dialog topic to
be an
independent, self-selected category superordinate to a co-constructed sequence
of dialog contributions
[8].
2.1
Topic Shifts
A dialog topic subordinates a sequence of coherent dialog contributions as wholes [9,2].
Hence, they generalize the concepts mentioned in these contributions to a certain de-
gree. A potential
topic shift
in dialogs occurs, once previous concepts and concepts
coming up subsequently cannot be generalized to one topic anymore. If attending to the
new concepts opens a completely different dialog topic and comes along with a drop of
the present one, we refer to this kind of shift as
topic leap
[2].
