Journal Articles
|
Schulz, Axel; Guckelsberger, Christian; Janssen, Frederik Semantic Abstraction for Generalization of Tweet Classification: An Evaluation on Incident-Related Tweets (Journal Article) In: Semantic Web, 2015. (Abstract | Links | BibTeX) @article{Schulz2015b,
title = {Semantic Abstraction for Generalization of Tweet Classification: An Evaluation on Incident-Related Tweets},
author = {Axel Schulz and Christian Guckelsberger and Frederik Janssen},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/03/guckelsberger_semantic_web_2015.pdf},
year = {2015},
date = {2015-01-01},
journal = {Semantic Web},
abstract = {Social media is a rich source of up-to-date information about events such as incidents. The sheer amount of available information makes machine learning approaches a necessity to process this information further. This learning problem is often concerned with regionally restricted datasets such as data from only one city. Because social media data such as tweets varies considerably across different cities, the training of efficient models requires labeling data from each city of interest, which is costly and time consuming. To avoid such an expensive labeling procedure, a generalizable model can be trained on data from one city and then applied to data from different cities. In this paper, we present Semantic Abstraction to improve the generalization of tweet classification. In particular, we derive features from Linked Open Data and include location and temporal mentions. A comprehensive evaluation on twenty datasets from ten different cities shows that Semantic Abstraction is indeed a valuable means for improving general- ization. We show that this not only holds for a two-class problem where incident-related tweets are separated from non-related ones but also for a four-class problem where three different incident types and a neutral class are distinguished. To get a thorough understanding of the generalization problem itself, we closely examined rule-based models from our evalu- ation. We conclude that on the one hand, the quality of the model strongly depends on the class distribution. On the other hand, the rules learned on cities with an equal class distribution are in most cases much more intuitive than those induced from skewed distributions. We also found that most of the learned rules rely on the novel semantically abstracted features.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Social media is a rich source of up-to-date information about events such as incidents. The sheer amount of available information makes machine learning approaches a necessity to process this information further. This learning problem is often concerned with regionally restricted datasets such as data from only one city. Because social media data such as tweets varies considerably across different cities, the training of efficient models requires labeling data from each city of interest, which is costly and time consuming. To avoid such an expensive labeling procedure, a generalizable model can be trained on data from one city and then applied to data from different cities. In this paper, we present Semantic Abstraction to improve the generalization of tweet classification. In particular, we derive features from Linked Open Data and include location and temporal mentions. A comprehensive evaluation on twenty datasets from ten different cities shows that Semantic Abstraction is indeed a valuable means for improving general- ization. We show that this not only holds for a two-class problem where incident-related tweets are separated from non-related ones but also for a four-class problem where three different incident types and a neutral class are distinguished. To get a thorough understanding of the generalization problem itself, we closely examined rule-based models from our evalu- ation. We conclude that on the one hand, the quality of the model strongly depends on the class distribution. On the other hand, the rules learned on cities with an equal class distribution are in most cases much more intuitive than those induced from skewed distributions. We also found that most of the learned rules rely on the novel semantically abstracted features. |
Guckelsberger, Christian; Polani, Daniel Effects of Anticipation in Individually Motivated Behaviour on Survival and Control in a Multi-Agent Scenario with Resource Constraints (Journal Article) In: Entropy, 16 (6), pp. 3357–3378, 2014, ISSN: 1099-4300. (Abstract | Links | BibTeX) @article{Guckelsberger2014,
title = {Effects of Anticipation in Individually Motivated Behaviour on Survival and Control in a Multi-Agent Scenario with Resource Constraints},
author = {Christian Guckelsberger and Daniel Polani},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/03/guckelsberger_gso13.pdf},
issn = {1099-4300},
year = {2014},
date = {2014-01-01},
journal = {Entropy},
volume = {16},
number = {6},
pages = {3357--3378},
abstract = {Self-organization and survival are inextricably bound to an agent's ability to control and anticipate its environment. Here we assess both skills when multiple agents compete for a scarce resource. Drawing on insights from psychology, microsociology and control theory, we examine how different assumptions about the behaviour of an agent's peers in the anticipation process affect subjective control and survival strategies. To quantify control and drive behaviour, we use the recently developed information-theoretic quantity of empowerment with the principle of empowerment maximization. In two experiments involving extensive simulations, we show that agents develop risk-seeking, risk-averse and mixed strategies, which correspond to greedy, parsimonious and mixed behaviour. Although the principle of empowerment maximization is highly generic, the emerging strategies are consistent with what one would expect from rational individuals with dedicated utility models. Our results support empowerment maximization as a universal drive for guided self-organization in collective agent systems.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Self-organization and survival are inextricably bound to an agent's ability to control and anticipate its environment. Here we assess both skills when multiple agents compete for a scarce resource. Drawing on insights from psychology, microsociology and control theory, we examine how different assumptions about the behaviour of an agent's peers in the anticipation process affect subjective control and survival strategies. To quantify control and drive behaviour, we use the recently developed information-theoretic quantity of empowerment with the principle of empowerment maximization. In two experiments involving extensive simulations, we show that agents develop risk-seeking, risk-averse and mixed strategies, which correspond to greedy, parsimonious and mixed behaviour. Although the principle of empowerment maximization is highly generic, the emerging strategies are consistent with what one would expect from rational individuals with dedicated utility models. Our results support empowerment maximization as a universal drive for guided self-organization in collective agent systems. |
Inproceedings
|
Biehl, Martin; Guckelsberger, Christian; Salge, Christoph; Smith, Simon; Polani, Daniel Free Energy, Empowerment, And Predictive Information Compared (Inproceedings) In: Proc. 9th Int. Conf. on Guided Self-Organisation (GSO'18), 2018. (Links | BibTeX) @inproceedings{Biehl2018a,
title = {Free Energy, Empowerment, And Predictive Information Compared},
author = {Martin Biehl and Christian Guckelsberger and Christoph Salge and Simon Smith and Daniel Polani},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2018/01/GSO_extended_abstract_Biehl_2018.pdf},
year = {2018},
date = {2018-01-01},
booktitle = {Proc. 9th Int. Conf. on Guided Self-Organisation (GSO'18)},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
|
Roohi, Shaghayegh; Takatalo, Jari; Guckelsberger, Christian; Hämäläinen, Perttu Review of Intrinsic Motivation in Simulation-based Game Testing (Inproceedings) In: Proc. 36th ACM Conf. Human Factors in Computing Systems (CHI'18), 2018. (Abstract | BibTeX) @inproceedings{Roohi2018,
title = {Review of Intrinsic Motivation in Simulation-based Game Testing},
author = {Shaghayegh Roohi and Jari Takatalo and Christian Guckelsberger and Perttu Hämäläinen},
year = {2018},
date = {2018-01-01},
booktitle = {Proc. 36th ACM Conf. Human Factors in Computing Systems (CHI'18)},
abstract = {This paper presents a review of intrinsic motivation in player modeling, with a focus on simulation-based game testing. Modern AI agents can learn to win many games; from a game testing perspective, a remaining research problem is how to model the aspects of human player behavior not explained by purely rational and goal-driven decision making. A major piece of this puzzle is constituted by intrinsic motivations, i.e., psychological needs that drive behavior without extrinsic reinforcement such as game score. We first review the common intrinsic motivations discussed in player psychology research and artificial intelligence, and then proceed to systematically review how the various motivations have been implemented in simulated player agents. Our work reveals that although motivations such as competence and curiosity have been studied in AI, work on utilizing them in simulation-based game testing is sparse, and other motivations such as social relatedness, immersion, and domination appear particularly underexplored.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper presents a review of intrinsic motivation in player modeling, with a focus on simulation-based game testing. Modern AI agents can learn to win many games; from a game testing perspective, a remaining research problem is how to model the aspects of human player behavior not explained by purely rational and goal-driven decision making. A major piece of this puzzle is constituted by intrinsic motivations, i.e., psychological needs that drive behavior without extrinsic reinforcement such as game score. We first review the common intrinsic motivations discussed in player psychology research and artificial intelligence, and then proceed to systematically review how the various motivations have been implemented in simulated player agents. Our work reveals that although motivations such as competence and curiosity have been studied in AI, work on utilizing them in simulation-based game testing is sparse, and other motivations such as social relatedness, immersion, and domination appear particularly underexplored. |
Guckelsberger, Christian; Salge, Christoph; Gow, Jeremy; Cairns, Paul Predicting Player Experience Without the Player. An Exploratory Study (Inproceedings) In: Proc. ACM Symp. on Computer-Human Interaction in Play (CHIPlay’17), 2017. (Abstract | Links | BibTeX) @inproceedings{GuckelsbergerCHIPlay2017,
title = {Predicting Player Experience Without the Player. An Exploratory Study},
author = {Christian Guckelsberger and Christoph Salge and Jeremy Gow and Paul Cairns},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2017/09/ChiPlay2017_PredictingPXWithoutThePlayer_CameraReady_v2.pdf},
year = {2017},
date = {2017-10-01},
booktitle = {Proc. ACM Symp. on Computer-Human Interaction in Play (CHIPlay’17)},
abstract = {A key challenge of procedural content generation (PCG) is to evoke a certain player experience (PX), when we have no direct control over the content which gives rise to that experience. We argue that neither the rigorous methods to assess PX in HCI, nor specialised methods in PCG are sufficient, because they rely on a human in the loop. We propose to address this shortcoming by means of computational models of intrinsic motivation and AI game-playing agents. We hypothesise that our approach could be used to automatically predict PX across games and content types without relying on a human player or designer. We conduct an exploratory study in level generation based on empowerment, a specific model of intrinsic motivation. Based on a thematic analysis, we find that empowerment can be used to create levels with qualitatively different PX. We relate the identified experiences to established theories of PX in HCI and game design, and discuss next steps.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A key challenge of procedural content generation (PCG) is to evoke a certain player experience (PX), when we have no direct control over the content which gives rise to that experience. We argue that neither the rigorous methods to assess PX in HCI, nor specialised methods in PCG are sufficient, because they rely on a human in the loop. We propose to address this shortcoming by means of computational models of intrinsic motivation and AI game-playing agents. We hypothesise that our approach could be used to automatically predict PX across games and content types without relying on a human player or designer. We conduct an exploratory study in level generation based on empowerment, a specific model of intrinsic motivation. Based on a thematic analysis, we find that empowerment can be used to create levels with qualitatively different PX. We relate the identified experiences to established theories of PX in HCI and game design, and discuss next steps. |
Guckelsberger, Christian; Salge, Christoph; Colton, Simon Addressing the "Why?" in Computational Creativity: A Non-Anthropocentric, Minimal Model of Intentional Creative Agency (Inproceedings) In: Proc. 8th Int. Conf. Computational Creativity, 2017
, 2017. (Abstract | Links | BibTeX) @inproceedings{Guckelsberger2017,
title = {Addressing the "Why?" in Computational Creativity: A Non-Anthropocentric, Minimal Model of Intentional Creative Agency},
author = {Christian Guckelsberger and Christoph Salge and Simon Colton},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2017/05/iccc2017_guckelsberger.pdf},
year = {2017},
date = {2017-05-09},
booktitle = {Proc. 8th Int. Conf. Computational Creativity, 2017
},
abstract = {Generally, computational creativity (CC) systems cannot explain why they are being creative, without ultimately referring back to the values and goals of their designer. Answering the why? would allow for the attribution of intentional agency, and likely lead to a stronger perception of creativity. Enactive artificial intelligence, a framework inspired by autopoietic enactive cognitive science, equips us with the necessary conditions for a value function to reflect a system's own intrinsic goals. We translate the framework's general claims to CC and ground a system's creative activity intrinsically in the maintenance of its identity. We relate to candidate computational principles to realise enactive artificial agents, thus laying the foundations for a minimal, non-anthropocentric model of intentional creative agency. We discuss first implications for the design and evaluation of CC, and address why human-level intentional creative agency is so hard to achieve. We ultimately propose a new research direction in CC, where intentional creative agency is addressed bottom up.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Generally, computational creativity (CC) systems cannot explain why they are being creative, without ultimately referring back to the values and goals of their designer. Answering the why? would allow for the attribution of intentional agency, and likely lead to a stronger perception of creativity. Enactive artificial intelligence, a framework inspired by autopoietic enactive cognitive science, equips us with the necessary conditions for a value function to reflect a system's own intrinsic goals. We translate the framework's general claims to CC and ground a system's creative activity intrinsically in the maintenance of its identity. We relate to candidate computational principles to realise enactive artificial agents, thus laying the foundations for a minimal, non-anthropocentric model of intentional creative agency. We discuss first implications for the design and evaluation of CC, and address why human-level intentional creative agency is so hard to achieve. We ultimately propose a new research direction in CC, where intentional creative agency is addressed bottom up. |
Denisova, Alena; Guckelsberger, Christian; Zendle, David Challenge in Digital Games: Towards Developing a Measurement Tool (Inproceedings) In: Proc. 35st ACM Conf. Human Factors in Computing Systems (CHI), ACM 2017. (Links | BibTeX) @inproceedings{Denisova2017,
title = {Challenge in Digital Games: Towards Developing a Measurement Tool},
author = {Alena Denisova and Christian Guckelsberger and David Zendle},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2017/05/denisova_chi2017_challenge.pdf},
year = {2017},
date = {2017-05-01},
booktitle = {Proc. 35st ACM Conf. Human Factors in Computing Systems (CHI)},
organization = {ACM},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
|
Guckelsberger, Christian; Salge, Christoph; Colton, Simon Intrinsically Motivated General Companion NPCs via Coupled Empowerment Maximisation (Inproceedings) In: Proc. IEEE Conf. Computational Intelligence in Games (CIG’16), IEEE, 2016. (Abstract | Links | BibTeX) @inproceedings{guckelsberger2016c,
title = {Intrinsically Motivated General Companion NPCs via Coupled Empowerment Maximisation},
author = {Christian Guckelsberger and Christoph Salge and Simon Colton},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/09/guckelsberger_cig16.pdf},
year = {2016},
date = {2016-09-01},
booktitle = {Proc. IEEE Conf. Computational Intelligence in Games (CIG’16)},
publisher = {IEEE},
abstract = {Non-player characters (NPCs) in games are traditionally hard-coded or dependent on pre-specified goals, and consequently struggle to behave sensibly in ever-changing and possibly unpredictable game worlds. To make them fit for new developments in procedural content generation, we introduce the principle of Coupled Empowerment Maximisation as an intrinsic motivation for game NPCs. We focus on the development of a general game companion, designed to support the player in achieving their goals. We evaluate our approach against three intuitive and abstract companion duties. We develop dedicated scenarios for each duty in a dungeon-crawler game testbed, and provide qualitative evidence that the emergent NPC behaviour fulfils these duties. We argue that this generic approach can speed up NPC AI development, improve automatic game evolution and introduce NPCs to full game-generation systems.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Non-player characters (NPCs) in games are traditionally hard-coded or dependent on pre-specified goals, and consequently struggle to behave sensibly in ever-changing and possibly unpredictable game worlds. To make them fit for new developments in procedural content generation, we introduce the principle of Coupled Empowerment Maximisation as an intrinsic motivation for game NPCs. We focus on the development of a general game companion, designed to support the player in achieving their goals. We evaluate our approach against three intuitive and abstract companion duties. We develop dedicated scenarios for each duty in a dungeon-crawler game testbed, and provide qualitative evidence that the emergent NPC behaviour fulfils these duties. We argue that this generic approach can speed up NPC AI development, improve automatic game evolution and introduce NPCs to full game-generation systems. |
Guckelsberger, Christian; Salge, Christoph Does Empowerment Maximisation Allow for Enactive Artificial Agents? (Inproceedings) In: Proc. 15th Int. Conf. Synthesis and Simulation of Living Systems (ALIFE), 2016. (Abstract | Links | BibTeX) @inproceedings{Guckelsberger2016,
title = {Does Empowerment Maximisation Allow for Enactive Artificial Agents?},
author = {Christian Guckelsberger and Christoph Salge},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/06/EmpowermentEnactiveAI_Submission2.pdf},
year = {2016},
date = {2016-07-01},
booktitle = {Proc. 15th Int. Conf. Synthesis and Simulation of Living Systems (ALIFE)},
abstract = {The enactive AI framework wants to overcome the sense making limitations of embodied AI by drawing on the biosystemic foundations of enactive cognitive science. While embodied AI tries to ground meaning in sensorimotor interaction, enactive AI adds further requirements by grounding sensorimotor interaction in autonomous agency. At the core of this shift is the requirement for a truly intrinsic value function. We suggest that empowerment, an information-theoretic quantity based on an agent’s embodiment, represents such a function. We highlight the role of empowerment maximization in satisfying the requirements of enactive AI, i.e. establishing constitutive autonomy and adaptivity, in detail. We then argue that empowerment, grounded in a precarious existence, allows an agent to enact a world based on the relevance of environmental features in respect to its own identity.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The enactive AI framework wants to overcome the sense making limitations of embodied AI by drawing on the biosystemic foundations of enactive cognitive science. While embodied AI tries to ground meaning in sensorimotor interaction, enactive AI adds further requirements by grounding sensorimotor interaction in autonomous agency. At the core of this shift is the requirement for a truly intrinsic value function. We suggest that empowerment, an information-theoretic quantity based on an agent’s embodiment, represents such a function. We highlight the role of empowerment maximization in satisfying the requirements of enactive AI, i.e. establishing constitutive autonomy and adaptivity, in detail. We then argue that empowerment, grounded in a precarious existence, allows an agent to enact a world based on the relevance of environmental features in respect to its own identity. |
Guckelsberger, Christian; Salge, Christoph; Saunders, Rob; Colton, Simon Supportive and Antagonistic Behaviour in Distributed Computational Creativity via Coupled Empowerment Maximisation (Inproceedings) In: Proc. 7th Int. Conf. Computational Creativity, 2016. (Abstract | Links | BibTeX) @inproceedings{Guckelsberger2016a,
title = {Supportive and Antagonistic Behaviour in Distributed Computational Creativity via Coupled Empowerment Maximisation},
author = {Christian Guckelsberger and Christoph Salge and Rob Saunders and Simon Colton},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/06/iccc_camera_ready.pdf},
year = {2016},
date = {2016-07-01},
booktitle = {Proc. 7th Int. Conf. Computational Creativity},
abstract = {There has been a strong tendency in distributed computational creativity systems to embrace embodied and situated agents for their flexible and adaptive behaviour. Intrinsically motivated agents are particularly successful in this respect, because they do not rely on externally specified goals, and can thus react flexibly to changes in open-ended environments. While supportive and antagonistic behaviour is omnipresent when people interact in creative tasks, existing implementations cannot establish such behaviour without constraining their agents’ flexibility by means of explicitly specified interaction rules. More open approaches in contrast cannot guarantee that support or antagonistic behaviour ever comes about. We define the information-theoretic principle of coupled empowerment maximisation as an intrinsically motivated frame for supportive and antagonistic behaviour within which agents can interact with maximum flexibility. We provide an intuition and a formalisation for an arbitrary number of agents. We then draw on several case-studies of co-creative and social creativity systems to make detailed predictions of the potential effect the underlying empowerment maximisation principle might have on the behaviour of creative agents.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
There has been a strong tendency in distributed computational creativity systems to embrace embodied and situated agents for their flexible and adaptive behaviour. Intrinsically motivated agents are particularly successful in this respect, because they do not rely on externally specified goals, and can thus react flexibly to changes in open-ended environments. While supportive and antagonistic behaviour is omnipresent when people interact in creative tasks, existing implementations cannot establish such behaviour without constraining their agents’ flexibility by means of explicitly specified interaction rules. More open approaches in contrast cannot guarantee that support or antagonistic behaviour ever comes about. We define the information-theoretic principle of coupled empowerment maximisation as an intrinsically motivated frame for supportive and antagonistic behaviour within which agents can interact with maximum flexibility. We provide an intuition and a formalisation for an arbitrary number of agents. We then draw on several case-studies of co-creative and social creativity systems to make detailed predictions of the potential effect the underlying empowerment maximisation principle might have on the behaviour of creative agents. |
Llano, Maria Teresa; Guckelsberger, Christian; Hepworth, Rose; Gow, Jeremy; Corneli, Joseph; Colton, Simon What If A Fish Got Drunk? Exploring the Plausibility of Machine-Generated Fictions (Inproceedings) In: Proc. 7th Int. Conf. Computational Creativity, 2016. (Abstract | Links | BibTeX) @inproceedings{Llano2016,
title = {What If A Fish Got Drunk? Exploring the Plausibility of Machine-Generated Fictions},
author = {Maria Teresa Llano and Christian Guckelsberger and Rose Hepworth and Jeremy Gow and Joseph Corneli and Simon Colton},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/06/ICCC_Plausibility_camera_ready-1.pdf},
year = {2016},
date = {2016-07-01},
booktitle = {Proc. 7th Int. Conf. Computational Creativity},
abstract = {Within the WHIM project, we study fictional ideation: processes for automatically inventing, assessing and presenting fictional ideas. Here we examine the foundational notion of the plausibility of fictional ideas, by performing an empirical study to surface the factors that affect judgements of plausibility. Our long term aim is to formalise a computational method which captures some intuitive notions of plausibility and can predict how certain types of people will assess the plausibility of certain types of fictional ideas. This paper constitutes a first firm step towards this aim.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Within the WHIM project, we study fictional ideation: processes for automatically inventing, assessing and presenting fictional ideas. Here we examine the foundational notion of the plausibility of fictional ideas, by performing an empirical study to surface the factors that affect judgements of plausibility. Our long term aim is to formalise a computational method which captures some intuitive notions of plausibility and can predict how certain types of people will assess the plausibility of certain types of fictional ideas. This paper constitutes a first firm step towards this aim. |
Schulz, Axel; Guckelsberger, Christian; Schmidt, Benedikt More Features Are Not Always Better: Evaluating Generalizing Models in Incident Type Classification of Tweets (Inproceedings) In: Proc. Conf. Empirical Methods in Natural Language Processing (EMNLP), 2015. (Abstract | Links | BibTeX) @inproceedings{Schulz2015a,
title = {More Features Are Not Always Better: Evaluating Generalizing Models in Incident Type Classification of Tweets},
author = {Axel Schulz and Christian Guckelsberger and Benedikt Schmidt},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/03/guckelsberger_emnlp_2015.pdf},
year = {2015},
date = {2015-01-01},
booktitle = {Proc. Conf. Empirical Methods in Natural Language Processing (EMNLP)},
abstract = {Social media represents a rich source of up-to-date information about events such as incidents. The sheer amount of available information makes machine learning approaches a necessity for further processing. This learn- ing problem is often concerned with region- ally restricted datasets such as data from only one city. Because social media data such as tweets varies considerably across different cities, the training of efficient models requires labeling data from each city of interest, which is costly and time consuming. In this study, we investigate which features are most suitable for training generalizable models, i.e., models that show good per- formance across different datasets. We re- implemented the most popular features from the state of the art in addition to other novel approaches, and evaluated them on data from ten different cities. We show that many so- phisticated features are not necessarily valuable for training a generalized model and are outperformed by classic features such as plain word-n-grams and character-n-grams.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Social media represents a rich source of up-to-date information about events such as incidents. The sheer amount of available information makes machine learning approaches a necessity for further processing. This learn- ing problem is often concerned with region- ally restricted datasets such as data from only one city. Because social media data such as tweets varies considerably across different cities, the training of efficient models requires labeling data from each city of interest, which is costly and time consuming. In this study, we investigate which features are most suitable for training generalizable models, i.e., models that show good per- formance across different datasets. We re- implemented the most popular features from the state of the art in addition to other novel approaches, and evaluated them on data from ten different cities. We show that many so- phisticated features are not necessarily valuable for training a generalized model and are outperformed by classic features such as plain word-n-grams and character-n-grams. |
Corneli, Joseph; Jordanous, Anna; Shepperd, Rosie; Llano, Maria Teresa; Misztal, Joanna; Colton, Simon; Guckelsberger, Christian Computational Poetry Workshop: Making Sense of Work in Progress (Inproceedings) In: Proc. 6th Int. Conf. Computational Creativity, 2015. (Abstract | Links | BibTeX) @inproceedings{Corneli2015,
title = {Computational Poetry Workshop: Making Sense of Work in Progress},
author = {Joseph Corneli and Anna Jordanous and Rosie Shepperd and Maria Teresa Llano and Joanna Misztal and Simon Colton and Christian Guckelsberger},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/03/corneli_iccc15_poetry-1.pdf},
year = {2015},
date = {2015-01-01},
booktitle = {Proc. 6th Int. Conf. Computational Creativity},
abstract = {Creativity cannot exist in a vacuum; it develops through feedback, learning, reflection and social interaction with others. However, this perspective has been relat- ively under-investigated in computational creativity re- search, which typically examines systems that operate individually. We develop a thought experiment showing how structured dialogues can help develop the creative aspects of computer poetry. Centrally in this approach, we ask questions of a poem, inviting it to tell us in what way it may be considered a “creative making.”},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Creativity cannot exist in a vacuum; it develops through feedback, learning, reflection and social interaction with others. However, this perspective has been relat- ively under-investigated in computational creativity re- search, which typically examines systems that operate individually. We develop a thought experiment showing how structured dialogues can help develop the creative aspects of computer poetry. Centrally in this approach, we ask questions of a poem, inviting it to tell us in what way it may be considered a “creative making.” |
Llano, Maria Teresa; Cook, Michael; Guckelsberger, Christian; Colton, Simon; Hepworth, Rose Towards the Automatic Generation of Fictional Ideas for Games (Inproceedings) In: Experimental AI in Games (EXAG’14), a workshop collocated with the tenth annual AAAI conference on artificial intelligence and interactive digital entertainment (AIIDE’14). AAAI Publications, 2014. (Abstract | Links | BibTeX) @inproceedings{llano2014towards,
title = {Towards the Automatic Generation of Fictional Ideas for Games},
author = { Maria Teresa Llano and Michael Cook and Christian Guckelsberger and Simon Colton and Rose Hepworth},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/09/llano_exag14.pdf},
year = {2014},
date = {2014-01-01},
booktitle = {Experimental AI in Games (EXAG’14), a workshop collocated with the tenth annual AAAI conference on artificial intelligence and interactive digital entertainment (AIIDE’14). AAAI Publications},
abstract = {The invention of fictional ideas is often a central pro- cess in the creative production of artefacts such as po- ems, music, paintings and games. Currently, fictional ideation is being studied by the Computational Creativ- ity community within the WHIM European project. The aim of WHIM is to develop the What-If Machine, a soft- ware system capable of inventing, evaluating and pre- senting fictional ideas with cultural value. In this pa- per we explore the potential applications of the What-If Machine in the context of games. Specifically, we pro- pose ways in which the What-If Machine can be used as an assistant for the design of games, by providing ideas about characters, the environment, etc., as well as a creative system during gameplay, through interesting interactions with the player.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The invention of fictional ideas is often a central pro- cess in the creative production of artefacts such as po- ems, music, paintings and games. Currently, fictional ideation is being studied by the Computational Creativ- ity community within the WHIM European project. The aim of WHIM is to develop the What-If Machine, a soft- ware system capable of inventing, evaluating and pre- senting fictional ideas with cultural value. In this pa- per we explore the potential applications of the What-If Machine in the context of games. Specifically, we pro- pose ways in which the What-If Machine can be used as an assistant for the design of games, by providing ideas about characters, the environment, etc., as well as a creative system during gameplay, through interesting interactions with the player. |
Miscellaneous
|
Corneli, Joe; Guckelsberger, Christian; Jordanous, Anna; Pease, Alison; Colton, Simon; Erden, Yasemin J Conference Report: AISB Members Workshop VII – Serendipity Symposium (Miscellaneous) AISB Quarterly (147), 2017. (Links | BibTeX) @misc{Corneli2017AISBQ,
title = {Conference Report: AISB Members Workshop VII – Serendipity Symposium},
author = {Joe Corneli and Christian Guckelsberger and Anna Jordanous and Alison Pease and Simon Colton and Yasemin J. Erden},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2017/09/serendipity_symposium_report_AISBQ.pdf},
year = {2017},
date = {2017-08-01},
howpublished = {AISB Quarterly (147)},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
|
Guckelsberger, Christian Conference Report: Eighth International Conference on Computational Creativity (Miscellaneous) AISB Quarterly (147), 2017. (Links | BibTeX) @misc{Guckelsberger2017AISBQ,
title = {Conference Report: Eighth International Conference on Computational Creativity},
author = {Christian Guckelsberger},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2017/09/ICCC_report_AISBQ.pdf},
year = {2017},
date = {2017-08-01},
howpublished = {AISB Quarterly (147)},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
|
Guckelsberger, Christian; Probst, Florian; Schulz, Axel Patent: Recommender System Employing Subjective Properties (pending) (Miscellaneous) US 20160132954 A1, 2016. (Abstract | Links | BibTeX) @misc{Guckelsberger2016b,
title = {Patent: Recommender System Employing Subjective Properties (pending)},
author = {Christian Guckelsberger and Florian Probst and Axel Schulz},
url = {https://www.google.com/patents/US20160132954?dq=guckelsberger&hl=en&sa=X&ved=0ahUKEwjvzM76qIDSAhVJAcAKHQ1hCaEQ6AEIKjAC},
year = {2016},
date = {2016-05-12},
abstract = {Example systems and methods of recommending an item are presented. In one example, preference values for items by multiple users, as well as property values for multiple properties of the items by the users, are accessed. Reference property values for the properties of the items are generated based on the property values. Average deviations from the reference property values for the properties across a first group of the items by a target user are generated. Expected property values for the properties of a second group of the items for the target user are generated based on the reference property values and the average deviations. Preference values of the target user for the second group of the items are estimated based on the expected property values. At least one of the second group of the items is recommended to the target user based on the estimated preference values.
},
howpublished = {US 20160132954 A1},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
Example systems and methods of recommending an item are presented. In one example, preference values for items by multiple users, as well as property values for multiple properties of the items by the users, are accessed. Reference property values for the properties of the items are generated based on the property values. Average deviations from the reference property values for the properties across a first group of the items by a target user are generated. Expected property values for the properties of a second group of the items for the target user are generated based on the reference property values and the average deviations. Preference values of the target user for the second group of the items are estimated based on the expected property values. At least one of the second group of the items is recommended to the target user based on the estimated preference values.
|
Grebner, Olaf; Bruchmann, Max; Guckelsberger, Christian; Probst, Florian; Schulz, Axel Patent: Reporting and Managing Incidents (Miscellaneous) US 8786433 B2, 2014. (Abstract | Links | BibTeX) @misc{Grebner2014,
title = {Patent: Reporting and Managing Incidents},
author = {Olaf Grebner and Max Bruchmann and Christian Guckelsberger and Florian Probst and Axel Schulz},
url = {https://scholar.google.co.uk/citations?view_op=view_citation&hl=en&user=tRV5TQkAAAAJ&citation_for_view=tRV5TQkAAAAJ:2osOgNQ5qMEC},
year = {2014},
date = {2014-07-22},
abstract = {Various embodiments of systems and methods for reporting and managing incidents are described herein. In one aspect, the method includes identifying a user logged into an incident reporting and alerting portal. A category for the logged-in user is identified. The category includes one of a reporter, a volunteer, and a responder. An incident report is received from the reporter. It is determined whether a comment is received on the incident report from the responder. When the comment is received, the comment is notified to the reporter. It is determined whether a reply to the comment is received from the reporter. When the reply to the comment is received, the reply is notified to the responder who commented on the incident report.},
howpublished = {US 8786433 B2},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
Various embodiments of systems and methods for reporting and managing incidents are described herein. In one aspect, the method includes identifying a user logged into an incident reporting and alerting portal. A category for the logged-in user is identified. The category includes one of a reporter, a volunteer, and a responder. An incident report is received from the reporter. It is determined whether a comment is received on the incident report from the responder. When the comment is received, the comment is notified to the reporter. It is determined whether a reply to the comment is received from the reporter. When the reply to the comment is received, the reply is notified to the responder who commented on the incident report. |
Technical Reports
|
Guckelsberger, Christian; Schulz, Axel STATSREP-ML: Statistical evaluation & reporting framework for machine learning results (Technical Report) Telecooperation Group, Technical University Darmstadt 2014. (Abstract | Links | BibTeX) @techreport{guckelsberger2014statsrep,
title = {STATSREP-ML: Statistical evaluation & reporting framework for machine learning results},
author = { Christian Guckelsberger and Axel Schulz},
url = {http://ccg.doc.gold.ac.uk/wp-content/uploads/2016/03/guckelsberger_statsrep_ml.pdf},
year = {2014},
date = {2014-01-01},
institution = {Telecooperation Group, Technical University Darmstadt},
abstract = {Data mining aims at finding regularities and patterns in data or extracting valuable data automatically by differentiating noise from relevant information. For example, in the emergency management domain, social media data is automatically filtered using information extraction and machine learning techniques [8, 6, 7]. During the last decades, many different classifiers and approaches for feature extraction have been developed to perform this automatic classification. However, finding the best approaches is challenging due to the large number of possible combinations.},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
Data mining aims at finding regularities and patterns in data or extracting valuable data automatically by differentiating noise from relevant information. For example, in the emergency management domain, social media data is automatically filtered using information extraction and machine learning techniques [8, 6, 7]. During the last decades, many different classifiers and approaches for feature extraction have been developed to perform this automatic classification. However, finding the best approaches is challenging due to the large number of possible combinations. |
