Detection of Question-Answer Pairs in Email Conversations
5808Authors
nan
Abstract
While sentence extraction as an approach to summarization has been shown to work in docu- ments of certain genres, because of the conver- sational nature of email communication where utterances are made in relation to one made previously, sentence extraction may not capture the necessary segments of dialogue that would make a summary coherent. In this paper, we present our work on the detection of question- answer pairs in an email conversation for the task of email summarization. We show that var- ious features based on the structure of email- threads can be used to improve upon lexical similarity of discourse segments for question- answer pairing. 1
Keywords
Full Content
Detection of Question-Answer Pairs in Email Conversations Lokesh Shrestha and Kathleen McKeown Columbia University Computer Science Deparment New York, NY 10027, USA, lokesh@cs.columbia.edu, kathy@cs.columbia.edu Abstract While sentence extraction as an approach to summarization has been shown to work in docu- ments of certain genres, because of the conver- sational nature of email communication where utterances are made in relation to one made previously, sentence extraction may not capture the necessary segments of dialogue that would make a summary coherent. In this paper, we present our work on the detection of question- answer pairs in an email conversation for the task of email summarization. We show that var- ious features based on the structure of email- threads can be used to improve upon lexical similarity of discourse segments for question- answer pairing. 1 Introduction In this paper, we discuss work on the detection of question and answer pairs in email threads, i.e., co- herent exchanges of email messages among several participants. Email is a written medium of asyn- chronous multi-party communication. This means that, as in face-to-face spoken dialog, the email thread as a whole is a collaborative effort with inter- action among the discourse participants. However, unlike spoken dialog, the discourse participants are not physically co-present, so that the written word is the only channel of communication. Furthermore, replies do not happen immediately, so that respon- ders need to take special precautions to identify rel- evant elements of the discourse context (for exam- ple, by quoting previous messages). Thus, email is a distinct linguistic genre that poses its own chal- lenges to summarization. With the increasing popularity of email as a means of communication, an increasing number of meetings are scheduled, events planned, issues re- solved, and questions answered through emails. As the number of emails in one’s mailbox increases, in- Regarding “acm home/bjarney”, on Apr 9, 2001, Muriel Danslop wrote: Two things: Can someone be responsible for the press releases for Stroustrup? Responding to this on Apr 10, 2001, Theresa Feng wrote: I think Phil, who is probably a better writer than most of us, is writing up something for dang and Dave to send out to various ACM chapters. Phil, we can just use that as our “press release”, right? In another subthread, on Apr 12, 2001, Kevin Danquoit wrote: Are you sending out upcoming events for this week? Figure 1: Sample summary obtained with sentence extraction formation from past conversations becomes increas- ingly inaccessible and difficult to manage. For ex- ample, a number of emails can be used in schedul- ing a meeting, and a search for information on the meeting may retrieve all the intermediate emails, thus hindering one’s access to the required informa- tion. Access to required information could be dra- matically improved by querying summaries of email conversations While summarization of email conversations seems a natural way to improve upon current methods of email management, research on email summarization is in early stages. Consider an example summary of a thread of email conver- sation …
Content truncated to 500 words
Labeled Text
While sentence extraction as an approach to summarization has been shown to work in documents of certain genres, because of the conversational nature of email communication where utterances are made in relation to one made previously, sentence extraction may not capture the necessary segments of dialogue that would make a summary coherent. In this paper, we present our work on the detection of questionanswer pairs in an email conversation for the task of email summarization. We show that various features based on the structure of emailthreads can be used to improve upon lexical similarity of discourse segments for questionanswer pairing. 1
Publication Details
- Paper ID:
- 5808
- Year:
- 2004
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan