Publications
Search
401.
Bruijnes, Merijn; Akker, Rieks; Hartholt, Arno; Heylen, Dirk
Virtual Suspect William Proceedings Article
In: Intelligent Virtual Agents, pp. 67–76, Springer, 2015.
@inproceedings{bruijnes_virtual_2015,
title = {Virtual Suspect William},
author = {Merijn Bruijnes and Rieks Akker and Arno Hartholt and Dirk Heylen},
url = {http://ict.usc.edu/pubs/Virtual%20Suspect%20William.pdf},
year = {2015},
date = {2015-08-01},
booktitle = {Intelligent Virtual Agents},
pages = {67–76},
publisher = {Springer},
abstract = {We evaluate an algorithm which computes the responses of an agent that plays the role of a suspect in simulations of police interrogations. The algorithm is based on a cognitive model - the response model - that is centred around keeping track of interpersonal relations. The model is parametrized in such a way that different personalities of the virtual suspect can be defined. In the evaluation we defined three different personalities and had participants guess the personality based on the responses the model provided in an interaction with the participant. We investigate what factors contributed to the ability of a virtual agent to show behaviour that was recognized by participants as belonging to a persona.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We evaluate an algorithm which computes the responses of an agent that plays the role of a suspect in simulations of police interrogations. The algorithm is based on a cognitive model - the response model - that is centred around keeping track of interpersonal relations. The model is parametrized in such a way that different personalities of the virtual suspect can be defined. In the evaluation we defined three different personalities and had participants guess the personality based on the responses the model provided in an interaction with the participant. We investigate what factors contributed to the ability of a virtual agent to show behaviour that was recognized by participants as belonging to a persona.
402.
Gratch, Jonathan; Hill, Susan; Morency, Louis-Philippe; Pynadath, David; Traum, David
Exploring the Implications of Virtual Human Research for Human-Robot Teams Proceedings Article
In: Virtual, Augmented and Mixed Reality, pp. 186–196, Springer International Publishing, Los Angeles, CA, 2015, ISBN: 978-3-319-21066-7 978-3-319-21067-4.
@inproceedings{gratch_exploring_2015,
title = {Exploring the Implications of Virtual Human Research for Human-Robot Teams},
author = {Jonathan Gratch and Susan Hill and Louis-Philippe Morency and David Pynadath and David Traum},
url = {http://ict.usc.edu/pubs/Exploring%20the%20Implications%20of%20Virtual%20Human%20Research%20for%20Human-Robot%20Teams.pdf},
doi = {10.1007/978-3-319-21067-4_20},
isbn = {978-3-319-21066-7 978-3-319-21067-4},
year = {2015},
date = {2015-08-01},
booktitle = {Virtual, Augmented and Mixed Reality},
volume = {9179},
pages = {186–196},
publisher = {Springer International Publishing},
address = {Los Angeles, CA},
abstract = {This article briefly explores potential synergies between the fields of virtual human and human-robot interaction research. We consider challenges in advancing the effectiveness of human-robot teams makes recommendations for enhancing this by facilitating synergies between robotics and virtual human research.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This article briefly explores potential synergies between the fields of virtual human and human-robot interaction research. We consider challenges in advancing the effectiveness of human-robot teams makes recommendations for enhancing this by facilitating synergies between robotics and virtual human research.
403.
Ustun, Volkan; Rosenbloom, Paul S.
Towards Adaptive, Interactive Virtual Humans in Sigma Proceedings Article
In: Intelligent Virtual Agents, pp. 98 –108, Springer, Delft, Netherlands, 2015, ISBN: 978-3-319-21995-0.
@inproceedings{ustun_towards_2015,
title = {Towards Adaptive, Interactive Virtual Humans in Sigma},
author = {Volkan Ustun and Paul S. Rosenbloom},
url = {http://ict.usc.edu/pubs/Towards%20Adaptive,%20Interactive%20Virtual%20Humans%20in%20Sigma.pdf},
doi = {10.1007/978-3-319-21996-7_10},
isbn = {978-3-319-21995-0},
year = {2015},
date = {2015-08-01},
booktitle = {Intelligent Virtual Agents},
volume = {9238},
pages = {98 –108},
publisher = {Springer},
address = {Delft, Netherlands},
abstract = {Sigma is a nascent cognitive architecture/system that combines concepts from graphical models with traditional symbolic architectures. Here an initial Sigma-based virtual human (VH) is introduced that combines probabilistic reasoning, rule-based decision-making, Theory of Mind, Simultaneous Localization and Mapping and reinforcement learning in a unified manner. This non-modular unification of diverse cognitive, robotic and VH capabilities provides an important first step towards fully adaptive and interactive VHs in Sigma.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Sigma is a nascent cognitive architecture/system that combines concepts from graphical models with traditional symbolic architectures. Here an initial Sigma-based virtual human (VH) is introduced that combines probabilistic reasoning, rule-based decision-making, Theory of Mind, Simultaneous Localization and Mapping and reinforcement learning in a unified manner. This non-modular unification of diverse cognitive, robotic and VH capabilities provides an important first step towards fully adaptive and interactive VHs in Sigma.
404.
Hoegen, Rens; Stratou, Giota; Lucas, Gale M.; Gratch, Jonathan
Comparing Behavior Towards Humans and Virtual Humans in a Social Dilemma Proceedings Article
In: Intelligent Virtual Agents, pp. 452–460, Springer International Publishing, Delft, Netherlands, 2015, ISBN: 978-3-319-21995-0 978-3-319-21996-7.
@inproceedings{hoegen_comparing_2015,
title = {Comparing Behavior Towards Humans and Virtual Humans in a Social Dilemma},
author = {Rens Hoegen and Giota Stratou and Gale M. Lucas and Jonathan Gratch},
url = {http://ict.usc.edu/pubs/Comparing%20Behavior%20Towards%20Humans%20and%20Virtual%20Humans%20in%20a%20Social%20Dilemma.pdf},
doi = {10.1007/978-3-319-21996-7 48},
isbn = {978-3-319-21995-0 978-3-319-21996-7},
year = {2015},
date = {2015-08-01},
booktitle = {Intelligent Virtual Agents},
volume = {9238},
pages = {452–460},
publisher = {Springer International Publishing},
address = {Delft, Netherlands},
abstract = {The difference of shown social behavior towards virtual humans and real humans has been subject to much research. Many of these studies compare virtual humans (VH) that are presented as either virtual agents controlled by a computer or as avatars controlled by real humans. In this study we directly compare VHs with real humans. Participants played an economic game against a computer-controlled VH or a visible human opponent. Decisions made throughout the game were logged, additionally participants’ faces were filmed during the study and analyzed with expression recognition software. The analysis of choices showed participants are far more willing to violate social norms with VHs: they are more willing to steal and less willing to forgive. Facial expressions show trends that suggest they are treating VHs less socially. The results highlight, that even in impoverished social interactions, VHs have a long way to go before they can evoke truly human-like responses.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The difference of shown social behavior towards virtual humans and real humans has been subject to much research. Many of these studies compare virtual humans (VH) that are presented as either virtual agents controlled by a computer or as avatars controlled by real humans. In this study we directly compare VHs with real humans. Participants played an economic game against a computer-controlled VH or a visible human opponent. Decisions made throughout the game were logged, additionally participants’ faces were filmed during the study and analyzed with expression recognition software. The analysis of choices showed participants are far more willing to violate social norms with VHs: they are more willing to steal and less willing to forgive. Facial expressions show trends that suggest they are treating VHs less socially. The results highlight, that even in impoverished social interactions, VHs have a long way to go before they can evoke truly human-like responses.
405.
Gratch, Jonathan; DeVault, David; Lucas, Gale M.; Marsella, Stacy
Negotiation as a Challenge Problem for Virtual Humans Proceedings Article
In: Brinkman, Willem-Paul; Broekens, Joost; Heylen, Dirk (Ed.): Intelligent Virtual Agents, pp. 201–215, Springer International Publishing, Delft, Netherlands, 2015, ISBN: 978-3-319-21995-0 978-3-319-21996-7.
@inproceedings{gratch_negotiation_2015,
title = {Negotiation as a Challenge Problem for Virtual Humans},
author = {Jonathan Gratch and David DeVault and Gale M. Lucas and Stacy Marsella},
editor = {Willem-Paul Brinkman and Joost Broekens and Dirk Heylen},
url = {http://ict.usc.edu/pubs/Negotiation%20as%20a%20Challenge%20Problem%20for%20Virtual%20Humans.pdf},
doi = {10.1007/978-3-319-21996-7_21},
isbn = {978-3-319-21995-0 978-3-319-21996-7},
year = {2015},
date = {2015-08-01},
booktitle = {Intelligent Virtual Agents},
volume = {9238},
pages = {201–215},
publisher = {Springer International Publishing},
address = {Delft, Netherlands},
abstract = {We argue for the importance of negotiation as a challenge problem for virtual human research, and introduce a virtual conversational agent that allows people to practice a wide range of negotiation skills. We describe the multi-issue bargaining task, which has become a de facto standard for teaching and research on negotiation in both the social and computer sciences. This task is popular as it allows scientists or instructors to create a variety of distinct situations that arise in real-life negotiations, simply by manipulating a small number of mathematical parameters. We describe the development of a virtual human that will allow students to practice the interpersonal skills they need to recognize and navigate these situations. An evaluation of an early wizard-controlled version of the system demonstrates the promise of this technology for teaching negotiation and supporting scientific research on social intelligence.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We argue for the importance of negotiation as a challenge problem for virtual human research, and introduce a virtual conversational agent that allows people to practice a wide range of negotiation skills. We describe the multi-issue bargaining task, which has become a de facto standard for teaching and research on negotiation in both the social and computer sciences. This task is popular as it allows scientists or instructors to create a variety of distinct situations that arise in real-life negotiations, simply by manipulating a small number of mathematical parameters. We describe the development of a virtual human that will allow students to practice the interpersonal skills they need to recognize and navigate these situations. An evaluation of an early wizard-controlled version of the system demonstrates the promise of this technology for teaching negotiation and supporting scientific research on social intelligence.
406.
Nazari, Zahra; Lucas, Gale M.; Gratch, Jonathan
Opponent Modeling for Virtual Human Negotiators Proceedings Article
In: Intelligent Virtual Agents, pp. 39–49, Springer International Publishing, Delft, Netherlands, 2015, ISBN: 978-3-319-21995-0 978-3-319-21996-7.
@inproceedings{nazari_opponent_2015,
title = {Opponent Modeling for Virtual Human Negotiators},
author = {Zahra Nazari and Gale M. Lucas and Jonathan Gratch},
url = {http://ict.usc.edu/pubs/Opponent%20Modeling%20for%20Virtual%20Human%20Negotiators.pdf},
doi = {10.1007/978-3-319-21996-7_4},
isbn = {978-3-319-21995-0 978-3-319-21996-7},
year = {2015},
date = {2015-08-01},
booktitle = {Intelligent Virtual Agents},
volume = {9238},
pages = {39–49},
publisher = {Springer International Publishing},
address = {Delft, Netherlands},
abstract = {Negotiation is a challenging domain for virtual human research. One aspect of this problem, known as opponent modeling, is discovering what the other party wants from the negotiation. Research in automated negotiation has yielded a number opponent modeling techniques but we show that these methods do not easily transfer to human-agent settings. We propose a more effective heuristic for inferring preferences both from a negotiator’s pattern of offers and verbal statements about their preferences. This method has the added advantage that it can detect negotiators that lie about their preferences. We discuss several ways the method can enhance the capabilities of a virtual human negotiator.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Negotiation is a challenging domain for virtual human research. One aspect of this problem, known as opponent modeling, is discovering what the other party wants from the negotiation. Research in automated negotiation has yielded a number opponent modeling techniques but we show that these methods do not easily transfer to human-agent settings. We propose a more effective heuristic for inferring preferences both from a negotiator’s pattern of offers and verbal statements about their preferences. This method has the added advantage that it can detect negotiators that lie about their preferences. We discuss several ways the method can enhance the capabilities of a virtual human negotiator.
407.
Kang, Sin-Hwa; Feng, Andrew; Leuski, Anton; Casas, Dan; Shapiro, Ari
Smart Mobile Virtual Humans: “Chat with Me!” Proceedings Article
In: Proceedings of the 15th International Conference on Intelligent Virtual Agents (IVA), pp. 475–478, Springer, Delft, Netherlands, 2015.
@inproceedings{kang_smart_2015,
title = {Smart Mobile Virtual Humans: “Chat with Me!”},
author = {Sin-Hwa Kang and Andrew Feng and Anton Leuski and Dan Casas and Ari Shapiro},
url = {http://ict.usc.edu/pubs/Smart%20Mobile%20Virtual%20Humans%20-%20Chat%20with%20Me.pdf},
doi = {10.1007/978-3-319-21996-7},
year = {2015},
date = {2015-08-01},
booktitle = {Proceedings of the 15th International Conference on Intelligent Virtual Agents (IVA)},
pages = {475–478},
publisher = {Springer},
address = {Delft, Netherlands},
abstract = {In this study, we are interested in exploring whether people would talk with 3D animated virtual humans using a smartphone for a longer amount of time as a sign of feeling rapport [5], compared to non-animated or audio-only characters in everyday life. Based on previous studies [2, 7, 10], users prefer animated characters in emotionally engaged interactions when the characters were displayed on mobile devices, yet in a lab setting. We aimed to reach a broad range of users outside of the lab in natural settings to investigate the potential of our virtual human on smartphones to facilitate casual, yet emotionally engaging conversation. We also found that the literature has not reached a consensus regarding the ideal gaze patterns for a virtual human, one thing researchers agree on is that inappropriate gaze could negatively impact conversations at times, even worse than receiving no visual feedback at all [1, 4]. Everyday life may bring the experience of awkwardness or uncomfortable sentiments in reaction to continuous mutual gaze. On the other hand, gaze aversion could also make a speaker think their partner is not listening. Our work further aims to address this question of what constitutes appropriate eye gaze in emotionally engaged interactions. We developed a 3D animated and chat-based virtual human which presented emotionally expressive nonverbal behaviors such as facial expressions, head gestures, gaze, and other upper body movements (see Figure 1). The virtual human displayed appropriate gaze that was either consisted of constant mutual gaze or gaze aversion based on a statistical model of saccadic eye movement [8] while listening. Both gaze patterns were accompanied by other forms of appropriate nonverbal feedback. To explore the question of optimal communicative medium, we distributed our virtual human application to users via an app store for Android-powered phones (i.e. Google Play Store) in order to target users who owned a smartphone and could use our application in various natural settings.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
In this study, we are interested in exploring whether people would talk with 3D animated virtual humans using a smartphone for a longer amount of time as a sign of feeling rapport [5], compared to non-animated or audio-only characters in everyday life. Based on previous studies [2, 7, 10], users prefer animated characters in emotionally engaged interactions when the characters were displayed on mobile devices, yet in a lab setting. We aimed to reach a broad range of users outside of the lab in natural settings to investigate the potential of our virtual human on smartphones to facilitate casual, yet emotionally engaging conversation. We also found that the literature has not reached a consensus regarding the ideal gaze patterns for a virtual human, one thing researchers agree on is that inappropriate gaze could negatively impact conversations at times, even worse than receiving no visual feedback at all [1, 4]. Everyday life may bring the experience of awkwardness or uncomfortable sentiments in reaction to continuous mutual gaze. On the other hand, gaze aversion could also make a speaker think their partner is not listening. Our work further aims to address this question of what constitutes appropriate eye gaze in emotionally engaged interactions. We developed a 3D animated and chat-based virtual human which presented emotionally expressive nonverbal behaviors such as facial expressions, head gestures, gaze, and other upper body movements (see Figure 1). The virtual human displayed appropriate gaze that was either consisted of constant mutual gaze or gaze aversion based on a statistical model of saccadic eye movement [8] while listening. Both gaze patterns were accompanied by other forms of appropriate nonverbal feedback. To explore the question of optimal communicative medium, we distributed our virtual human application to users via an app store for Android-powered phones (i.e. Google Play Store) in order to target users who owned a smartphone and could use our application in various natural settings.
408.
Melo, Celso; Gratch, Jonathan
Beyond Believability: Quantifying the Differences Between Real and Virtual Humans Proceedings Article
In: Proceedings of IVA 2015, pp. 109 – 118, Springer, Delft, Netherlands, 2015.
@inproceedings{de_melo_beyond_2015,
title = {Beyond Believability: Quantifying the Differences Between Real and Virtual Humans},
author = {Celso Melo and Jonathan Gratch},
url = {http://ict.usc.edu/pubs/Beyond%20Believability%20-%20Quantifying%20the%20Differences%20between%20Real%20and%20Virtual%20Humans.pdf},
year = {2015},
date = {2015-08-01},
booktitle = {Proceedings of IVA 2015},
pages = {109 – 118},
publisher = {Springer},
address = {Delft, Netherlands},
abstract = {“Believable” agents are supposed to “suspend the audience’s disbelief” and provide the “illusion of life”. However, beyond such high-level definitions, which are prone to subjective interpretation, there is not much more to help researchers systematically create or assess whether their agents are believable. In this paper we propose a more pragmatic and useful benchmark than believability for designing virtual agents. This benchmark requires people, in a specific social situation, to act with the virtual agent in the same manner as they would with a real human. We propose that perceptions of mind in virtual agents, especially pertaining to agency – the ability to act and plan – and experience – the ability to sense and feel emotion – are critical for achieving this new benchmark. We also review current computational systems that fail, pass, and even surpass this benchmark and show how a theoretical framework based on perceptions of mind can shed light into these systems. We also discuss a few important cases where it is better if virtual humans do not pass the benchmark. We discuss implications for the design of virtual agents that can be as natural and efficient to interact with as real humans.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
“Believable” agents are supposed to “suspend the audience’s disbelief” and provide the “illusion of life”. However, beyond such high-level definitions, which are prone to subjective interpretation, there is not much more to help researchers systematically create or assess whether their agents are believable. In this paper we propose a more pragmatic and useful benchmark than believability for designing virtual agents. This benchmark requires people, in a specific social situation, to act with the virtual agent in the same manner as they would with a real human. We propose that perceptions of mind in virtual agents, especially pertaining to agency – the ability to act and plan – and experience – the ability to sense and feel emotion – are critical for achieving this new benchmark. We also review current computational systems that fail, pass, and even surpass this benchmark and show how a theoretical framework based on perceptions of mind can shed light into these systems. We also discuss a few important cases where it is better if virtual humans do not pass the benchmark. We discuss implications for the design of virtual agents that can be as natural and efficient to interact with as real humans.
409.
Rosenbloom, Paul S.; Gratch, Jonathan; Ustun, Volkan
Towards Emotion in Sigma: From Appraisal to Attention Proceedings Article
In: Proceedings of AGI 2015, pp. 142 – 151, Springer International Publishing, Berlin, Germany, 2015.
@inproceedings{rosenbloom_towards_2015,
title = {Towards Emotion in Sigma: From Appraisal to Attention},
author = {Paul S. Rosenbloom and Jonathan Gratch and Volkan Ustun},
url = {http://ict.usc.edu/pubs/Towards%20Emotion%20in%20Sigma%20-%20From%20Appraisal%20to%20Attention.pdf},
year = {2015},
date = {2015-07-01},
booktitle = {Proceedings of AGI 2015},
volume = {9205},
pages = {142 – 151},
publisher = {Springer International Publishing},
address = {Berlin, Germany},
abstract = {A first step is taken towards incorporating emotional processing into Sigma, a cognitive architecture that is grounded in graphical models, with the addition of appraisal variables for expectedness and desirability plus their initial implications for attention at two levels of the control hierarchy. The results leverage many of Sigma's existing capabilities but with a few key additions.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A first step is taken towards incorporating emotional processing into Sigma, a cognitive architecture that is grounded in graphical models, with the addition of appraisal variables for expectedness and desirability plus their initial implications for attention at two levels of the control hierarchy. The results leverage many of Sigma's existing capabilities but with a few key additions.
410.
Nouri, Elnaz; Traum, David
Cross cultural report of values and decisions in the multi round ultimatum game and the centipede game Proceedings Article
In: Proceeding of AHFE 2015, Las Vegas, NV, 2015.
@inproceedings{nouri_cross_2015,
title = {Cross cultural report of values and decisions in the multi round ultimatum game and the centipede game},
author = {Elnaz Nouri and David Traum},
url = {http://ict.usc.edu/pubs/Cross%20cultural%20report%20of%20values%20and%20decisions%20in%20the%20multi%20round%20ultimatum%20game%20and%20the%20centipede%20game.pdf},
year = {2015},
date = {2015-07-01},
booktitle = {Proceeding of AHFE 2015},
address = {Las Vegas, NV},
abstract = {This paper investigates the cultural differences in decision making behavior of people from the US and India. We study players from these cultures playing the Multi Round Ultimatum Game and the Centipede Game online. In order to study how people from different cultures evaluate decisions we use criteria from the Multi Attribute Relational Values (MARV) survey. Our results confirm the existence of cultural differences in how people from US and India make decisions in the Ultimatum and Centipede games. We also observe differences in responses to survey questions implying differences in the amount of importance that the two cultures assign to the MARV decision making criteria.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper investigates the cultural differences in decision making behavior of people from the US and India. We study players from these cultures playing the Multi Round Ultimatum Game and the Centipede Game online. In order to study how people from different cultures evaluate decisions we use criteria from the Multi Attribute Relational Values (MARV) survey. Our results confirm the existence of cultural differences in how people from US and India make decisions in the Ultimatum and Centipede games. We also observe differences in responses to survey questions implying differences in the amount of importance that the two cultures assign to the MARV decision making criteria.
411.
Gratch, Jon; Lucas, Gale; Malandrakis, Nikolaos; Szablowski, Evan; Fessler, Eli
To tweet or not to tweet: The question of emotion and excitement about sporting events Proceedings Article
In: Proceedings of the Bi-Annual Conference of the International Society for Research on Emotion, Geneva, Switzerland, 2015.
@inproceedings{gratch_tweet_2015,
title = {To tweet or not to tweet: The question of emotion and excitement about sporting events},
author = {Jon Gratch and Gale Lucas and Nikolaos Malandrakis and Evan Szablowski and Eli Fessler},
url = {http://ict.usc.edu/pubs/To%20tweet%20or%20not%20to%20tweet%20-The%20question%20of%20emotion%20and%20excitement%20about%20sporting%20events.pdf},
year = {2015},
date = {2015-07-01},
booktitle = {Proceedings of the Bi-Annual Conference of the International Society for Research on Emotion},
address = {Geneva, Switzerland},
abstract = {Sporting events can serve as laboratories to explore emotion and computational tools provide new ways to examine emotional processes “in the wild”. Moreover, emotional processes are assumed -but untested- in sports economics. For example, according to the well-studied uncertainty of outcome hypothesis (UOH), “close” games are more exciting and therefore better attended. If one team were certain to win, it would take away a major source of excitement, reducing positive affect, and therefore decreasing attendance. The role of emotion here is assumed but has not been tested; furthermore, the measures used (ticket sales, attendance, TV-viewership) do not allow for such a test because they are devoid of emotional content. To address this problem, we use tweets per minute (specifically, tweets posted during 2014 World Cup with official game hashtags). Sentiment analysis of these tweets can give interesting insights into what emotional processes are involved. Another benefit of tweets is that they are dynamic, and novel results from dynamic analyses (of TV-viewership) suggest that the UOH effect can actually reverse as games unfold (people switch channels away from close games). We therefore also reconsider the UOH, specifically, extending it by both examining sentiment and dynamic changes during the game. To consider such changes, we focus on games that could have been close (high in uncertainty), but ended up being lower in uncertainty. We operationalize such unexpected certainty of outcome as the extent to which games are predicted to be “close” (based on betting odds), but ended up with a bigger difference between the teams’ scores than was expected. Statistical analyses revealed that, contrary to the UOH, games with a bigger difference in score between teams than expected had higher tweets per minute. We also performed sentiment analysis, categorizing each tweet as positive, negative or neutral, and found that games with higher tweets per minute also have a higher percentage of negative tweets. Furthermore, games that have a bigger difference than expected have a higher percentage of negative tweets (compared to games closer to what is expected). This analysis seems to suggest that, contrary to assumptions in sports economics, excitement relates to expressions of negative emotion (and not positive emotion). The results are discussed in terms of innovations in methodology and understanding the role of emotion for “tuning in” to real world events. Further research could explore the specific mechanisms that link negative sentiment to excitement, such as worry or out-group derogation.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Sporting events can serve as laboratories to explore emotion and computational tools provide new ways to examine emotional processes “in the wild”. Moreover, emotional processes are assumed -but untested- in sports economics. For example, according to the well-studied uncertainty of outcome hypothesis (UOH), “close” games are more exciting and therefore better attended. If one team were certain to win, it would take away a major source of excitement, reducing positive affect, and therefore decreasing attendance. The role of emotion here is assumed but has not been tested; furthermore, the measures used (ticket sales, attendance, TV-viewership) do not allow for such a test because they are devoid of emotional content. To address this problem, we use tweets per minute (specifically, tweets posted during 2014 World Cup with official game hashtags). Sentiment analysis of these tweets can give interesting insights into what emotional processes are involved. Another benefit of tweets is that they are dynamic, and novel results from dynamic analyses (of TV-viewership) suggest that the UOH effect can actually reverse as games unfold (people switch channels away from close games). We therefore also reconsider the UOH, specifically, extending it by both examining sentiment and dynamic changes during the game. To consider such changes, we focus on games that could have been close (high in uncertainty), but ended up being lower in uncertainty. We operationalize such unexpected certainty of outcome as the extent to which games are predicted to be “close” (based on betting odds), but ended up with a bigger difference between the teams’ scores than was expected. Statistical analyses revealed that, contrary to the UOH, games with a bigger difference in score between teams than expected had higher tweets per minute. We also performed sentiment analysis, categorizing each tweet as positive, negative or neutral, and found that games with higher tweets per minute also have a higher percentage of negative tweets. Furthermore, games that have a bigger difference than expected have a higher percentage of negative tweets (compared to games closer to what is expected). This analysis seems to suggest that, contrary to assumptions in sports economics, excitement relates to expressions of negative emotion (and not positive emotion). The results are discussed in terms of innovations in methodology and understanding the role of emotion for “tuning in” to real world events. Further research could explore the specific mechanisms that link negative sentiment to excitement, such as worry or out-group derogation.
412.
Demski, Abram
Expression Graphs Unifying Factor Graphs and Sum-Product Networks Proceedings Article
In: Artificial General Intelligence, pp. 241–250, Springer, Berlin, Germany, 2015.
@inproceedings{demski_expression_2015,
title = {Expression Graphs Unifying Factor Graphs and Sum-Product Networks},
author = {Abram Demski},
url = {http://ict.usc.edu/pubs/Expression%20Graphs%20Unifying%20Factor%20Graphs%20and%20Sum-Product%20Networks.pdf},
year = {2015},
date = {2015-07-01},
booktitle = {Artificial General Intelligence},
pages = {241–250},
publisher = {Springer},
address = {Berlin, Germany},
abstract = {Factor graphs are a very general knowledge representation, subsuming many existing formalisms in AI. Sum-product networks are a more recent representation, inspired by studying cases where factor graphs are tractable. Factor graphs emphasize expressive power, while sum-product networks restrict expressiveness to get strong guarantees on speed of inference. A sum-product network is not simply a restricted factor graph, however. Although the inference algorithms for the two structures are very similar, translating a sum-product network into factor graph representation can result in an exponential slowdown. We propose a formalism which generalizes factor graphs and sum-product networks, such that inference is fast in cases whose structure is close to a sum-product network.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Factor graphs are a very general knowledge representation, subsuming many existing formalisms in AI. Sum-product networks are a more recent representation, inspired by studying cases where factor graphs are tractable. Factor graphs emphasize expressive power, while sum-product networks restrict expressiveness to get strong guarantees on speed of inference. A sum-product network is not simply a restricted factor graph, however. Although the inference algorithms for the two structures are very similar, translating a sum-product network into factor graph representation can result in an exponential slowdown. We propose a formalism which generalizes factor graphs and sum-product networks, such that inference is fast in cases whose structure is close to a sum-product network.
413.
Kommers, Cody; Ustun, Volkan; Demski, Abram; Rosenbloom, Paul
Hierarchical Reasoning with Distributed Vector Representations Proceedings Article
In: Proceedings of 37th Annual Conference of the Cognitive Science Society, Cognitive Science Society, Pasadena, CA, 2015.
@inproceedings{kommers_hierarchical_2015,
title = {Hierarchical Reasoning with Distributed Vector Representations},
author = {Cody Kommers and Volkan Ustun and Abram Demski and Paul Rosenbloom},
url = {http://ict.usc.edu/pubs/Hierarchical%20Reasoning%20with%20Distributed%20Vector%20Representations.pdf},
year = {2015},
date = {2015-07-01},
booktitle = {Proceedings of 37th Annual Conference of the Cognitive Science Society},
publisher = {Cognitive Science Society},
address = {Pasadena, CA},
abstract = {We demonstrate that distributed vector representations are capable of hierarchical reasoning by summing sets of vectors representing hyponyms (subordinate concepts) to yield a vector that resembles the associated hypernym (superordinate concept). These distributed vector representations constitute a potentially neurally plausible model while demonstrating a high level of performance in many different cognitive tasks. Experiments were run using DVRS, a word embedding system designed for the Sigma cognitive architecture, and Word2Vec, a state-of-the-art word embedding system. These results contribute to a growing body of work demonstrating the various tasks on which distributed vector representations perform competently.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We demonstrate that distributed vector representations are capable of hierarchical reasoning by summing sets of vectors representing hyponyms (subordinate concepts) to yield a vector that resembles the associated hypernym (superordinate concept). These distributed vector representations constitute a potentially neurally plausible model while demonstrating a high level of performance in many different cognitive tasks. Experiments were run using DVRS, a word embedding system designed for the Sigma cognitive architecture, and Word2Vec, a state-of-the-art word embedding system. These results contribute to a growing body of work demonstrating the various tasks on which distributed vector representations perform competently.
414.
Cummins, Nicholas; Scherer, Stefan; Krajewski, Jarek; Schnieder, Sebastian; Epps, Julien; Quatieri, Thomas F.
A Review of Depression and Suicide Risk Assessment Using Speech Analysis Journal Article
In: Speech Communication, vol. 71, pp. 10 – 49, 2015, ISSN: 0167-6393.
@article{cummins_review_2015,
title = {A Review of Depression and Suicide Risk Assessment Using Speech Analysis},
author = {Nicholas Cummins and Stefan Scherer and Jarek Krajewski and Sebastian Schnieder and Julien Epps and Thomas F. Quatieri},
url = {http://www.sciencedirect.com/science/article/pii/S0167639315000369},
doi = {http://dx.doi.org/10.1016/j.specom.2015.03.004},
issn = {0167-6393},
year = {2015},
date = {2015-07-01},
journal = {Speech Communication},
volume = {71},
pages = {10 – 49},
abstract = {This paper is the first review into the automatic analysis of speech for use as an objective predictor of depression and suicidality. Both conditions are major public health concerns; depression has long been recognised as a prominent cause of disability and burden worldwide, whilst suicide is a misunderstood and complex course of death that strongly impacts the quality of life and mental health of the families and communities left behind. Despite this prevalence the diagnosis of depression and assessment of suicide risk, due to their complex clinical characterisations, are difficult tasks, nominally achieved by the categorical assessment of a set of specific symptoms. However many of the key symptoms of either condition, such as altered mood and motivation, are not physical in nature; therefore assigning a categorical score to them introduces a range of subjective biases to the diagnostic procedure. Due to these difficulties, research into finding a set of biological, physiological and behavioural markers to aid clinical assessment is gaining in popularity. This review starts by building the case for speech to be considered a key objective marker for both conditions; reviewing current diagnostic and assessment methods for depression and suicidality including key non-speech biological, physiological and behavioural markers and highlighting the expected cognitive and physiological changes associated with both conditions which affect speech production. We then review the key characteristics; size, associated clinical scores and collection paradigm, of active depressed and suicidal speech databases. The main focus of this paper is on how common paralinguistic speech characteristics are affected by depression and suicidality and the application of this information in classification and prediction systems. The paper concludes with an in-depth discussion on the key challenges – improving the generalisability through greater research collaboration and increased standardisation of data collection, and the mitigating unwanted sources of variability – that will shape the future research directions of this rapidly growing field of speech processing research.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper is the first review into the automatic analysis of speech for use as an objective predictor of depression and suicidality. Both conditions are major public health concerns; depression has long been recognised as a prominent cause of disability and burden worldwide, whilst suicide is a misunderstood and complex course of death that strongly impacts the quality of life and mental health of the families and communities left behind. Despite this prevalence the diagnosis of depression and assessment of suicide risk, due to their complex clinical characterisations, are difficult tasks, nominally achieved by the categorical assessment of a set of specific symptoms. However many of the key symptoms of either condition, such as altered mood and motivation, are not physical in nature; therefore assigning a categorical score to them introduces a range of subjective biases to the diagnostic procedure. Due to these difficulties, research into finding a set of biological, physiological and behavioural markers to aid clinical assessment is gaining in popularity. This review starts by building the case for speech to be considered a key objective marker for both conditions; reviewing current diagnostic and assessment methods for depression and suicidality including key non-speech biological, physiological and behavioural markers and highlighting the expected cognitive and physiological changes associated with both conditions which affect speech production. We then review the key characteristics; size, associated clinical scores and collection paradigm, of active depressed and suicidal speech databases. The main focus of this paper is on how common paralinguistic speech characteristics are affected by depression and suicidality and the application of this information in classification and prediction systems. The paper concludes with an in-depth discussion on the key challenges – improving the generalisability through greater research collaboration and increased standardisation of data collection, and the mitigating unwanted sources of variability – that will shape the future research directions of this rapidly growing field of speech processing research.
415.
Chatterjee, Moitreya; Leuski, Anton
CRMActive: An Active Learning Based Approach for Effective Video Annotation and Retrieval Proceedings Article
In: Proceedings of ACM International Conference on Multimedia Retrieval (ICMR), pp. 535–538, ACM, Shanghai, China, 2015.
@inproceedings{chatterjee_crmactive_2015,
title = {CRMActive: An Active Learning Based Approach for Effective Video Annotation and Retrieval},
author = {Moitreya Chatterjee and Anton Leuski},
url = {http://ict.usc.edu/pubs/CRMActive%20-%20An%20Active%20Learning%20Based%20Approach%20for%20Effective%20Video%20Annotation%20and%20Retrieval.pdf},
doi = {10.1145/2671188.2749342},
year = {2015},
date = {2015-06-01},
booktitle = {Proceedings of ACM International Conference on Multimedia Retrieval (ICMR)},
pages = {535–538},
publisher = {ACM},
address = {Shanghai, China},
abstract = {Conventional multimedia annotation/retrieval systems such as Normalized Continuous Relevance Model (NormCRM) [7]require a fully labeled training data for a good performance. Active Learning, by determining an order for labeling the training data, allows for a good performance even before the training data is fully annotated. In this work we propose an active learning algorithm, which combines a novel measure of sample uncertainty with a novel clustering-based approach for determining sample density and diversity and integrate it with NormCRM. The clusters are also iteratively re⬚ned to ensure both feature and label-level agreement among samples. We show that our approach outperforms multiple baselines both on a new, open dataset and on the popular TRECVID corpus at both the tasks of annotation and text-based retrieval of videos.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Conventional multimedia annotation/retrieval systems such as Normalized Continuous Relevance Model (NormCRM) [7]require a fully labeled training data for a good performance. Active Learning, by determining an order for labeling the training data, allows for a good performance even before the training data is fully annotated. In this work we propose an active learning algorithm, which combines a novel measure of sample uncertainty with a novel clustering-based approach for determining sample density and diversity and integrate it with NormCRM. The clusters are also iteratively re⬚ned to ensure both feature and label-level agreement among samples. We show that our approach outperforms multiple baselines both on a new, open dataset and on the popular TRECVID corpus at both the tasks of annotation and text-based retrieval of videos.
416.
Scherer, Stefan; Lucas, Gale; Gratch, Jonathan; Rizzo, Albert; Morency, Louis-Philippe
Self-reported symptoms of depression and PTSD are associated with reduced vowel space in screening interviews Journal Article
In: IEEE Transactions on Affective Computing (in press; doi: 10.1109/TAFFC.2015.2440264), no. 99, 2015, ISSN: 1949-3045.
@article{scherer_self-reported_2015,
title = {Self-reported symptoms of depression and PTSD are associated with reduced vowel space in screening interviews},
author = {Stefan Scherer and Gale Lucas and Jonathan Gratch and Albert Rizzo and Louis-Philippe Morency},
url = {http://ict.usc.edu/pubs/Self-reported%20symptoms%20of%20depression%20and%20PTSD%20are%20associated%20with%20reduced%20vowel%20space%20in%20screening%20interviews.pdf},
doi = {10.1109/TAFFC.2015.2440264},
issn = {1949-3045},
year = {2015},
date = {2015-06-01},
journal = {IEEE Transactions on Affective Computing (in press; doi: 10.1109/TAFFC.2015.2440264)},
number = {99},
abstract = {Reduced frequency range in vowel production is a well documented speech characteristic of individuals with psychological and neurological disorders. Affective disorders such as depression and post-traumatic stress disorder (PTSD) are known to influence motor control and in particular speech production. The assessment and documentation of reduced vowel space and reduced expressivity often either rely on subjective assessments or on analysis of speech under constrained laboratory conditions (e.g. sustained vowel production, reading tasks). These constraints render the analysis of such measures expensive and impractical. Within this work, we investigate an automatic unsupervised machine learning based approach to assess a speaker’s vowel space. Our experiments are based on recordings of 253 individuals. Symptoms of depression and PTSD are assessed using standard self-assessment questionnaires and their cut-off scores. The experiments show a significantly reduced vowel space in subjects that scored positively on the questionnaires. We show the measure’s statistical robustness against varying demographics of individuals and articulation rate. The reduced vowel space for subjects with symptoms of depression can be explained by the common condition of psychomotor retardation influencing articulation and motor control. These findings could potentially support treatment of affective disorders, like depression and PTSD in the future.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Reduced frequency range in vowel production is a well documented speech characteristic of individuals with psychological and neurological disorders. Affective disorders such as depression and post-traumatic stress disorder (PTSD) are known to influence motor control and in particular speech production. The assessment and documentation of reduced vowel space and reduced expressivity often either rely on subjective assessments or on analysis of speech under constrained laboratory conditions (e.g. sustained vowel production, reading tasks). These constraints render the analysis of such measures expensive and impractical. Within this work, we investigate an automatic unsupervised machine learning based approach to assess a speaker’s vowel space. Our experiments are based on recordings of 253 individuals. Symptoms of depression and PTSD are assessed using standard self-assessment questionnaires and their cut-off scores. The experiments show a significantly reduced vowel space in subjects that scored positively on the questionnaires. We show the measure’s statistical robustness against varying demographics of individuals and articulation rate. The reduced vowel space for subjects with symptoms of depression can be explained by the common condition of psychomotor retardation influencing articulation and motor control. These findings could potentially support treatment of affective disorders, like depression and PTSD in the future.
417.
Knowles, Megan; Lucas, Gale; Baumeister, Roy; Gardner, Wendi
Choking Under Social Pressure: Social Monitoring Among the Lonely Journal Article
In: Personality and Social Psychology Bulletin, vol. 41, no. 6, pp. 805–821, 2015, ISSN: 0146-1672, 1552-7433.
@article{knowles_choking_2015,
title = {Choking Under Social Pressure: Social Monitoring Among the Lonely},
author = {Megan Knowles and Gale Lucas and Roy Baumeister and Wendi Gardner},
url = {http://ict.usc.edu/pubs/Choking%20Under%20Social%20Pressure%20-%20Social%20Monitoring%20Among%20the%20Lonely.pdf},
doi = {10.1177/0146167215580775},
issn = {0146-1672, 1552-7433},
year = {2015},
date = {2015-06-01},
journal = {Personality and Social Psychology Bulletin},
volume = {41},
number = {6},
pages = {805–821},
abstract = {Lonely individuals may decode social cues well but have difficulty putting such skills to use precisely when they need them—in social situations. In four studies, we examined whether lonely people choke under social pressure by asking participants to complete social sensitivity tasks framed as diagnostic of social skills or nonsocial skills. Across studies, lonely participants performed worse than nonlonely participants on social sensitivity tasks framed as tests of social aptitude, but they performed just as well or better than the nonlonely when the same tasks were framed as tests of academic aptitude. Mediational analyses in Study 3 and misattribution effects in Study 4 indicate that anxiety plays an important role in this choking effect. This research suggests that lonely individuals may not need to acquire social skills to escape loneliness; instead, they must learn to cope with performance anxiety in interpersonal interactions.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Lonely individuals may decode social cues well but have difficulty putting such skills to use precisely when they need them—in social situations. In four studies, we examined whether lonely people choke under social pressure by asking participants to complete social sensitivity tasks framed as diagnostic of social skills or nonsocial skills. Across studies, lonely participants performed worse than nonlonely participants on social sensitivity tasks framed as tests of social aptitude, but they performed just as well or better than the nonlonely when the same tasks were framed as tests of academic aptitude. Mediational analyses in Study 3 and misattribution effects in Study 4 indicate that anxiety plays an important role in this choking effect. This research suggests that lonely individuals may not need to acquire social skills to escape loneliness; instead, they must learn to cope with performance anxiety in interpersonal interactions.
418.
Artstein, Ron; Leuski, Anton; Maio, Heather; Mor-Barak, Tomer; Gordon, Carla; Traum, David
How Many Utterances Are Needed to Support Time-Offset Interaction? Proceedings Article
In: Proceedings of FLAIRS 28, pp. 144–149, AAAI Press, Hollywood, FL, 2015, ISBN: 978-1-57735-730-8.
@inproceedings{artstein_how_2015,
title = {How Many Utterances Are Needed to Support Time-Offset Interaction?},
author = {Ron Artstein and Anton Leuski and Heather Maio and Tomer Mor-Barak and Carla Gordon and David Traum},
url = {http://www.aaai.org/ocs/index.php/FLAIRS/FLAIRS15/paper/view/10442},
isbn = {978-1-57735-730-8},
year = {2015},
date = {2015-05-01},
booktitle = {Proceedings of FLAIRS 28},
pages = {144–149},
publisher = {AAAI Press},
address = {Hollywood, FL},
abstract = {A set of several hundred recorded statements by a single speaker is sufficient to address unrestricted questions and sustain short conversations on a circumscribed topic. Statements were recorded by Pinchas Gutter, a Holocaust survivor, talking about his personal experiences before, during and after the Holocaust. These statements were delivered to participants in conversation, using a “Wizard of Oz” system, where live operators select an appropriate reaction to each user utterance in real time. Even though participants were completely unconstrained in the questions they could ask, the recorded statements were able to directly address at least 58% of user questions. The unanswered questions were then analyzed to identify gaps, and additional statements were recorded to fill the gaps. The statements will be put in an automated system using existing language understanding technology, to create the first full working system of time-offset interaction, allowing a live conversation with a real human who is not present for the conversation in real time.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A set of several hundred recorded statements by a single speaker is sufficient to address unrestricted questions and sustain short conversations on a circumscribed topic. Statements were recorded by Pinchas Gutter, a Holocaust survivor, talking about his personal experiences before, during and after the Holocaust. These statements were delivered to participants in conversation, using a “Wizard of Oz” system, where live operators select an appropriate reaction to each user utterance in real time. Even though participants were completely unconstrained in the questions they could ask, the recorded statements were able to directly address at least 58% of user questions. The unanswered questions were then analyzed to identify gaps, and additional statements were recorded to fill the gaps. The statements will be put in an automated system using existing language understanding technology, to create the first full working system of time-offset interaction, allowing a live conversation with a real human who is not present for the conversation in real time.
419.
Lucas, Gale M.; McCubbins, Mathew D.; Turner, Mark
Against Game Theory Book Section
In: Emerging Trends in the Social and Behavioral Sciences: An Interdisciplinary, Searchable, and Linkable Resource, pp. 1–16, John Wiley & Sons, Inc., Hoboken, NJ, 2015, ISBN: 978-1-118-90077-2.
@incollection{lucas_against_2015,
title = {Against Game Theory},
author = {Gale M. Lucas and Mathew D. McCubbins and Mark Turner},
url = {http://ict.usc.edu/pubs/Against%20GameTheory.pdf},
isbn = {978-1-118-90077-2},
year = {2015},
date = {2015-05-01},
booktitle = {Emerging Trends in the Social and Behavioral Sciences: An Interdisciplinary, Searchable, and Linkable Resource},
pages = {1–16},
publisher = {John Wiley & Sons, Inc.},
address = {Hoboken, NJ},
abstract = {People make choices. Often, the outcome depends on choices other people make. What mental steps do people go through when making such choices? Game theory, the most influential model of choice in economics and the social sciences, offers an answer, one based on games of strategy like chess and checkers: the chooser considers the choices that others will make and makes a choice that will lead to a better outcome for the chooser, given all those choices by other people. It is universally established in the social sciences that classical game theory (even when heavily modified) is bad at predicting behavior. But instead of abandoning classical game theory, those in the social sciences have mounted a rescue operation under the name of “behavioral game theory.” Its main tool is to propose systematic deviations from the predictions of game theory, deviations that arise from character type, for example. Other deviations purportedly come from cognitive overload or limitations. The fundamental idea of behavioral game theory is that, if we know the deviations, then we can correct our predictions accordingly, and so get it right. There are two problems with this rescue operation, each of them fatal. (1) For a chooser, contemplating the range of possible deviations, as there are many dozens, actually makes it exponentially harder to figure out a path to an outcome. This makes the theoretical models useless for modeling human thought or human behavior in general. (2) Modeling deviations is helpful only if the deviations are consistent, so that scientists (and indeed decision-makers) can make 2 predictions about future choices on the basis of past choices. But the deviations are not consistent. In general, deviations from classical models are not consistent for any individual from one task to the next or between individuals for the same task. In addition, people’s beliefs are in general not consistent with their choices. Accordingly, all hope is hollow that we can construct a general behavioral game theory. What can replace it? We survey some of the emerging candidates.},
keywords = {},
pubstate = {published},
tppubtype = {incollection}
}
People make choices. Often, the outcome depends on choices other people make. What mental steps do people go through when making such choices? Game theory, the most influential model of choice in economics and the social sciences, offers an answer, one based on games of strategy like chess and checkers: the chooser considers the choices that others will make and makes a choice that will lead to a better outcome for the chooser, given all those choices by other people. It is universally established in the social sciences that classical game theory (even when heavily modified) is bad at predicting behavior. But instead of abandoning classical game theory, those in the social sciences have mounted a rescue operation under the name of “behavioral game theory.” Its main tool is to propose systematic deviations from the predictions of game theory, deviations that arise from character type, for example. Other deviations purportedly come from cognitive overload or limitations. The fundamental idea of behavioral game theory is that, if we know the deviations, then we can correct our predictions accordingly, and so get it right. There are two problems with this rescue operation, each of them fatal. (1) For a chooser, contemplating the range of possible deviations, as there are many dozens, actually makes it exponentially harder to figure out a path to an outcome. This makes the theoretical models useless for modeling human thought or human behavior in general. (2) Modeling deviations is helpful only if the deviations are consistent, so that scientists (and indeed decision-makers) can make 2 predictions about future choices on the basis of past choices. But the deviations are not consistent. In general, deviations from classical models are not consistent for any individual from one task to the next or between individuals for the same task. In addition, people’s beliefs are in general not consistent with their choices. Accordingly, all hope is hollow that we can construct a general behavioral game theory. What can replace it? We survey some of the emerging candidates.
420.
Mell, Johnathan; Lucas, Gale; Gratch, Jonathan
An Effective Conversation Tactic for Creating Value over Repeated Negotiations Proceedings Article
In: Proceedings of the 2015 International Conference on Autonomous Agents and Multiagent Systems, pp. 1567–1576, International Foundation for Autonomous Agents and Multiagent Systems, Istanbul, Turkey, 2015.
@inproceedings{mell_effective_2015,
title = {An Effective Conversation Tactic for Creating Value over Repeated Negotiations},
author = {Johnathan Mell and Gale Lucas and Jonathan Gratch},
url = {http://ict.usc.edu/pubs/An%20Effective%20Conversation%20Tactic%20for%20Creating%20Value%20over%20Repeated%20Negotiations.pdf},
year = {2015},
date = {2015-05-01},
booktitle = {Proceedings of the 2015 International Conference on Autonomous Agents and Multiagent Systems},
pages = {1567–1576},
publisher = {International Foundation for Autonomous Agents and Multiagent Systems},
address = {Istanbul, Turkey},
abstract = {Automated negotiation research focuses on getting the most value from a single negotiation, yet real-world settings often involve repeated serial negotiations between the same parties. Repeated negotiations are interesting because they allow the discovery of mutually beneficial solutions that don’t exist within the confines of a single negotiation. This paper introduces the notion of Pareto efficiency over time to formalize this notion of value-creation through repeated interactions. We review literature from human negotiation research and identify a dialog strategy, favors and ledgers, that facilitates this process. As part of a longer-term effort to build intelligent virtual humans that can train human negotiators, we create a conversational agent that instantiates this strategy, and assess its effectiveness with human users, using the established Colored Trails negotiation testbed. In an empirical study involving a series of repeated negotiations, we show that humans are more likely to discover Pareto optimal solutions overtime when matched with our favor-seeking agent. Further, an agent that asks for favors during early negotiations, regardless of whether these favors are ever repaid, leads participants to discover more joint value in later negotiations, even under the traditional definition of Pareto optimality within a single negotiation. Further, agents that match their words with deeds (repay their favors) create the most value for themselves. We discuss the implications of these findings for agents that engage in long-term interactions with human users.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Automated negotiation research focuses on getting the most value from a single negotiation, yet real-world settings often involve repeated serial negotiations between the same parties. Repeated negotiations are interesting because they allow the discovery of mutually beneficial solutions that don’t exist within the confines of a single negotiation. This paper introduces the notion of Pareto efficiency over time to formalize this notion of value-creation through repeated interactions. We review literature from human negotiation research and identify a dialog strategy, favors and ledgers, that facilitates this process. As part of a longer-term effort to build intelligent virtual humans that can train human negotiators, we create a conversational agent that instantiates this strategy, and assess its effectiveness with human users, using the established Colored Trails negotiation testbed. In an empirical study involving a series of repeated negotiations, we show that humans are more likely to discover Pareto optimal solutions overtime when matched with our favor-seeking agent. Further, an agent that asks for favors during early negotiations, regardless of whether these favors are ever repaid, leads participants to discover more joint value in later negotiations, even under the traditional definition of Pareto optimality within a single negotiation. Further, agents that match their words with deeds (repay their favors) create the most value for themselves. We discuss the implications of these findings for agents that engage in long-term interactions with human users.
Filter
Sorry, no publications matched your criteria.