Information Technology Reference
In-Depth Information
Enabling a coherent dialog between Max and a human user necessitates the pre-
setting of a topical structure allowing to introduce the emulated topic awareness into
the dialog and to cause a corresponding conversational behavior on the agent's part.
Schneider (1988) assigns a structure to a typical small talk sequence as follows:
1. Question
2. Answer
3. Reply
4. Further turns
Furthermore, a typical small talk topically covers the immediated, external, and com-
munication situation [11]. In their study, Endrass et al. (2011) identified a typical dis-
tribution of these topics within a dialog between Germans. Thus, Germans address less
of the immediated and approximately equivalent of the external and communication
situation during small talk. According to these findings, and considering the conditions
arising from the fact that Max is situated in an university environment, the beginning of
topical small talks with the conversational agent Max is structured as follows: In his first
turn, Max asks the interlocutor for his or her subject of study as most potential dialog
partners are students. Subsequently, the agent tries to find out the interlocutor's origin.
If successful, Max is able to determine the interlocutor's favorite football club from this
knowledge and to continue with the football topic. In case the human does not want to
talk about football or sports in general, he or she has the opportunity to suggest another
topic. Thus, the first dialog topics are solely initiated by Max. This is important insofar
as this scenario also establishes a basis for the development of a personal memory for
the agent as introduced by Mattar and Wachsmuth [19]. This requires the gathering of
a lot of information concerning social categories about the human interlocutor.
5.1
Planned Evaluation
Upon successfully completing a preliminary evaluation identifying the topics of news-
paper articles, which has shown an average accuracy of 61.0 [8], we plan to accomplish
a more adequate evaluation considering and addressing the introduced dialog scenario.
Accordingly, we searched for a corpus comprising dialog information of German small
talks occurring during first encounters between two persons. The CUBE-G corpora [4]
provides analyzed records of 21 first interaction scenarios each between a student and
a professional actor and each lasting around five minutes. Amongst others, the dialogs
were tested for the amount of topics and topic shifts which is why the corpus contains
topical annotations for each recorded small talk. Thus, the CUBE-G corpus presents the
optimal basis for our following evaluation.
In preparation for the planned evaluation, we already determined a list of predefined
main categories that represent typical dialog topics for everyday small talks. Thereby,
we omitted so-called
unsafe topics
(see Section 1) and especially focused on topics
raised in the given university scenario. Table 1 shows the resulting list of main cate-
gories. Moreover, we downloaded the German database dump from May 14, 2011 and
built up a Lucene index containing all information parts relevant for our purpose.
