Politeness and Bias in Dialogue Summarization: Two Exploratory Studies

Abstract

In this chapter, two empirical pilot studies on the role of politeness in dialogue summarization are described. In these studies, a collection of four dialogues was used. Each dialogue was automatically generated by the NECA system and the politeness of the dialogue participants was systematically manipulated. Subjects were divided into groups who had to summarize the dialogues from a particular dialogue participant's point of view or the point of view of an impartial observer. In the first study, there were no other constraints. In the second study, the summarizers were restricted to summaries whose length did not exceed 10% of the number of words in the dialogue that was being summarized.Amongst other things, it was found that the politeness of the interaction is included more often in summaries of dialogues that deviate from what would be considered normal or unmarked. A comparison of the results of the two studies suggests that the extent to which politeness is reported is not affected by how long a summary is allowed to be. It was also found that the point of view of the summarizer influences which information is included in the summary and how it is presented. This finding did not seem to be affected by the constraint in our second study on the summary length.

Analyzing collaborative learning processes automatically: Exploiting the advances of computational linguistics in computer-supported collaborative learning

“…While typical applications of summarization technology are largely oriented towards extracting the most contentful sentences from a document, or collecting the most contentful sentences across multiple documents reporting about the same event (Kupiec et al 1995;Carbonell and Goldstein 1998;Gong and Liu 2001), conversation summarization is different. Behavioral studies about dialogue summarization show that what people consider important to include in a summary about a dialogue may include aspects of the nature of the conversation in addition to a condensed version of the information that was communicated (Roman et al 2006). Insights gained from process analysis of conversation logs from collaborative learning studies could enable the construction of summaries to support group moderators who do not have time to follow all of the details of every conversation occurring in parallel in an on-line learning environment.…”

Using Summarization to Discover Argument Facets in Online Idealogical Dialog

“…Other work examines how social phenomena affect summarization, such as a study of how the politeness level in computer-generated dialogs impacted summaries (Roman et al, 2006). Emotion naturally occurs in the IAC, and summarizers' orientation to emotion is intriguing.…”

Summarizing Dialogic Arguments from Social Media

“…(Zechner, 2001;Murray et al, 2006;Whittaker et al, 2012;Janin et al, 2004;Carletta, 2007). In contrast to information content, Roman et al (2006) examine how social phenomena such as politeness level affect summarization. Emotional information has also been observed in summaries of professional chats discussing technology (Zhou and Hovy, 2005).…”