Browse Prior Art Database

System for detecting the optimal point of conversation interruption using contextual input in a business teleconferencing environment

IP.com Disclosure Number: IPCOM000247341D
Publication Date: 2016-Aug-25
Document File: 4 page(s) / 153K

Publishing Venue

The IP.com Prior Art Database

Abstract

Automatically defining the optimal point in conversation for addressing someone (whether in a multiparty conversation, or in a dialogue) during a teleconference meeting is a difficult task as it implies different contextual knowledge. While in face-to-face meeting, a wide variety of inputs allow the participants to identify the most appropriate moment for taking the floor, in teleconferenced meetings, conversations can stumble due to participants talking over one-another, lengthy pauses, and missed cues for conversational inputs. In this publication we propose a system for automatically detecting the optimal interruption point in a multiparty teleconference meeting. The system is based on the analysis of the meeting's conversation flow.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

System for detecting the optimal point of conversation interruption using contextual input in a business teleconferencing environment

Introduction

Automatically defining the optimal point in conversation for addressing someone (whether in a multiparty conversation, or in a dialogue) during a teleconference meeting is a difficult task as it implies different contextual knowledge. While in face-to-face meeting, a wide variety of inputs allow the participants to determine the most appropriate moment for taking the floor, in teleconferenced meetings, conversations can stumble due to participants talking over one-another, lengthy pauses, and missed cues for conversational inputs.

    Our approach proposes a system for automatically detecting optimal interruption point in a teleconferenced meeting. Background

    There is a sizable amount of research into the areas of interruption and interruption management from a variety of different contexts and environments, from computers to dialogue systems for in-car communication systems. Much of the prior art in this area is research papers focused to study the dialogue flow and the turn taking dynamics in bi- and multiparty conversations.

Table 1 summarizes some of the prior art in the area.

    To the best of our knowledge, no work has been tailored on using these studies to detect the best moment for interruption in teleconference environment.

Article

Difference

The paper focused on the specific domain of car dialogue systems, and proposes strategy for changing topic, interrupting and resuming a conversation in this specific context. Our approach covers the domain of telefconferenced meetings, tackles different problems and uses different solutions. http://www.ling.gu.se/~jessi

ca/papers/gotal_villing_et_

al.pdf

Interruption, resumption and domain-switching in-vehicle dialogue [4]

Tracking Initiative in Collaborative Dialogue Interactions [3]

Paper focused on tracking

who will take the initiative

in the next dialogue turn.


Page 02 of 4

http://www.aclweb.org/anth ology/P/P97/P97-1034.pdf

Mobile Intelligent Interruption Management System [1]

Solution regards a device for automated administration of personal unavailability with regard to cell phone. The system is able to detect good moments for the cellular to interrupt. However, the focus of this solution is handling interruption with the purpose of re-handling requests of the user's attention and at a later convenient time, and within the confines of mobile/cellular input-outputs for a single person/device. Also, the device is focused on delaying calls at appropriate time (that do not interrupt users' actions); it does not consider the interruption issue within a conversation. http://www.jucs.org/jucs_1 6_15/a_mobile_intelligent_

interruption/jucs_16_15_2 060_2080_zulkernain.pdf

Solution

    The core idea of the solution dis...