Your SlideShare is downloading. ×
0
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Agent-Oriented Systems: From the Primitive to the Emotional
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Agent-Oriented Systems: From the Primitive to the Emotional

222

Published on

Talk given at University of Valencia July 25, 2013

Talk given at University of Valencia July 25, 2013

Published in: Technology, Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
222
On Slideshare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Agent-Oriented Systems: From the Primitive to the Emotional Dr. Aladdin Ayesh De Montfort University Talk given at University of Valencia July 25, 2013 Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 2. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Snapshot about De Montfort University - Faculty of Technology De Montfort University (DMU) is based in Leicester, which is in the East Midlands region of England. Leicester is about an hour by fast train from London and sets between Oxford, Cambridge and Nottingham. DMU consists of four major faculties: Faculty of Technology Faculty of Art, Design and Humanities Faculty of Business and Law Health and Life Sciences Faculty of Technology consists of two schools: School of Computer Science and Informatics School of Engineering, Media and Sustainable Development Agents and MAS related research is spread between a number of research centres in the faculty including STRL, CCI and IESD. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 3. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Agenda 1 Agents and Notion of Agency Notion of Agency Agents as structures and architectures Context 2 Oscillating Emotions Modelling Emotions Emotional Swarms 3 Applied Projects eLearning Mobility Text Mining 4 Collaboration with Valencia University Emotions in eLearning 5 Conclusion Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 4. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Notion of Agency: The Principle Agent technologies have emerged from Artificial Intelligence and Social Sciences studies to become an emergent filed within the main stream of Software Engineering. Yet these technologies are based on the principle notion of Agency. So what does this mean? To simplify the explanation let us take any ”agent” type and examine it: Estate agent (individual) Travel agent (individual) Commercial or distribution agent (company) Chemical agent (substance) What all these different agents have in common? Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 5. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Notion of Agency: me, myself and I The notion of agency is encompassed in the individual (the agent) and it’s abilities to identify itself, consciously or unconsciously, in relation to it’s self, surrounding and others. This lead us to some of the principles of agents and agent-oriented systems: interaction autonomy internal representations behavioural expressions Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 6. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 7. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 8. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 9. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 10. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust * Easily Maintainable Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 11. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust * Easily Maintainable * Portable Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 12. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust * Easily Maintainable * Portable * Adaptable and can easily be Updated Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 13. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust * Easily Maintainable * Portable * Adaptable and can easily be Updated This makes agent-based systems in compliance with most of software engineering desirable system features. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 14. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures I From software engineering viewpoint, agents are interesting structures that can be used to build systems with: * Dynamic architectures * Scalable * Robust * Easily Maintainable * Portable * Adaptable and can easily be Updated This makes agent-based systems in compliance with most of software engineering desirable system features. But agents give us more ... Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 15. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures II The flexibility of the agent structure and plasticity of agent’s definition, allow us to define an agent as: Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 16. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures II The flexibility of the agent structure and plasticity of agent’s definition, allow us to define an agent as: - A simple data structure such as a mindless particle that is represented by Position and Velocity < P, V > Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 17. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures II The flexibility of the agent structure and plasticity of agent’s definition, allow us to define an agent as: - A simple data structure such as a mindless particle that is represented by Position and Velocity < P, V > - An encompassing structure like classes in OOP with the addition of a communication interface, such is the case with most of software agents. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 18. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures II The flexibility of the agent structure and plasticity of agent’s definition, allow us to define an agent as: - A simple data structure such as a mindless particle that is represented by Position and Velocity < P, V > - An encompassing structure like classes in OOP with the addition of a communication interface, such is the case with most of software agents. - An all encompassing system with a degree of autonomy and communication and interaction protocols, such is the case with cognitive agents. The complexity of the system can vary and so does the level of intelligence. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 19. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures III In other words, agents can go from That exhibits intelligence only in a collective to that requires a psychologist! Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 20. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures III In other words, agents can go from The Primitive That exhibits intelligence only in a collective to that requires a psychologist! Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 21. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Agents as structures and architectures III In other words, agents can go from The Primitive That exhibits intelligence only in a collective to The Emotional that requires a psychologist! Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 22. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Platforms There are several platforms for agents-based system development. - Jade - Repast-Simphony - Netlogo - Agentlets - Jason - FIPA-OS Two good sources of information on platforms and agents standards are FIPA and EU AgentLink. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 23. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Concepts and Terminology Some basic concepts and terminologies that distinguish agent-oriented systems. - Delegation - Responsibility - Social Norms - Trust Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 24. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Notion of Agency Agents as structures and architectures Context Context and Limitation Agents-oriented systems field is large and wide in scope. It is by nature a multi-disciplinary subject criss-crossing multiple areas. Thus this talk, whilst giving a birds-eye view of agent technologies, will focus on specific component of an agent, that is The Representational System In particular on representing Emotions. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 25. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Modelling Emotions Emotions work as regulators of human behaviours. They influence the actions selection process. In some cases, they act as mere preference, which often view as a reflection of the personality of the actor. In other cases, emotions are too complicated to understand in separation of the full cognitive architecture of the human mind. Modelling emotions oscillate between these two facts even in psychology of emotions one finds several theories from simple set of basic emotions to emotional spaces. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 26. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotions Models I The simplest of emotion models are models based on Darwain’s basic emotions: Happiness, Sadness, Anger, Fear, Surprise, Disgust. The interpretation of these emotions into computational models vary from basic thresholds to a set of inference rules. Another set of computational models of emotions are based on single emotion theories, in particular stress. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 27. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotions Models II We worked with three psychological theories of emotions: We examined these models in reinforcement learning, game character controller and facial expression analysis. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 28. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotions Models II We worked with three psychological theories of emotions: 1 Darwin’s Basic Emotions using both discreet and fuzzy set representation. We examined these models in reinforcement learning, game character controller and facial expression analysis. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 29. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotions Models II We worked with three psychological theories of emotions: 1 Darwin’s Basic Emotions using both discreet and fuzzy set representation. 2 Milenson 3 dimensional space of emotions using fuzzy sets type I and II, and swarm-based emotional model. We examined these models in reinforcement learning, game character controller and facial expression analysis. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 30. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotions Models II We worked with three psychological theories of emotions: 1 Darwin’s Basic Emotions using both discreet and fuzzy set representation. 2 Milenson 3 dimensional space of emotions using fuzzy sets type I and II, and swarm-based emotional model. 3 Geneva wheel inspired model of emotions using fuzzy sets type I and II We examined these models in reinforcement learning, game character controller and facial expression analysis. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 31. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotional Swarms I One of key applications of agent-based swarm intelligence is crowd simulation. The simplest form of crowd is flocks of birds, thus one of the earliest and most successful swarm intelligence algorithms is flocking algorithm. In simulating human crowd, however, flocking-like algorithms give us limited scope for realistic simulation. Injecting crow elements with aspects of emotional models to influence their action selection, produces interesting results, e.g. panic as an emergent behaviour rather than coded. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 32. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotional Swarms II The greatest use of swarm algorithms, which sometimes exceed their original use in simulations, is optimization. Several machine learning algorithms were produced for this purpose: Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 33. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotional Swarms II The greatest use of swarm algorithms, which sometimes exceed their original use in simulations, is optimization. Several machine learning algorithms were produced for this purpose: 1 Particle Swarm Optimization (PSO) is very popular with engineering applications and in most cases provide an alternative to GAs. The swarm element here is the mindless particle. Agent technologies have little or nothing to do with it. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 34. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Emotional Swarms II The greatest use of swarm algorithms, which sometimes exceed their original use in simulations, is optimization. Several machine learning algorithms were produced for this purpose: 1 Particle Swarm Optimization (PSO) is very popular with engineering applications and in most cases provide an alternative to GAs. The swarm element here is the mindless particle. Agent technologies have little or nothing to do with it. 2 Ant Colony Optimization (ACO) is an advanced variation on Ant System and is very popular with solving optimization problems that can be presented as graphs, e.g. networks. Agent technologies play core role in developing these algorithms. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 35. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Modelling Emotions Emotional Swarms Social Cognition There are variations on insects and other biological agents like swarms. The more complex the agents involved the more complex is the interaction and the role that sensing-feeling-emotions regulators can play such is the case in reinforcement learning. This introduces the subject of social cognition which is at the heart of many applications of crowd simulation and it is applications in business intelligence, building design, urban development, etc. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 36. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining eLearning I eLearning systems are by nature distributed and dynamic systems which are perfect for the use of agents technologies. We used agents to manage the infrastructure of eLearning systems on data girds. We used them to manage the learning process itself to enable student mobility between institutes and course programs. But as the use of serious gaming and mobile devices is expanding there are several venues where agents can be used. Such as ... Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 37. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining eLearning II - Agents, coupled with gaming technologies, can be avatars for the learner in virtual and distributed class rooms. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 38. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining eLearning II - Agents, coupled with gaming technologies, can be avatars for the learner in virtual and distributed class rooms. - Agents, coupled with preference models, can be the engines behind interactive HCI to enable the learning process. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 39. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining eLearning II - Agents, coupled with gaming technologies, can be avatars for the learner in virtual and distributed class rooms. - Agents, coupled with preference models, can be the engines behind interactive HCI to enable the learning process. - Agents, coupled with policy languages, NLP, gaming and emotions, can be virtual managers of the user interaction with an eLearning system. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 40. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining Mobility I Mobile devices are spreading at a rapid speed not seen with any other technology. To enable these devices a host of hardware and software infrastructures were developed, continue to be developed and need continuous maintenance and management. Agent technologies provide solutions at the various levels of mobility from managing network infrastructure, to services run on the network, to services provided to the end user. As an example for location-based services ... Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 41. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining Mobility II With the introduction of easily accessible smartphones, Cloud Computing, 4G mobile networks, Web 2.0, and Social Media technologies, data volumes and availability are increasing exponentially. Dynamic scalable system architectures and new protocols to implement these systems are ever more needed. Equally, systems that exploit this rich stream of data are also in demands. Thus ... Content management * Sentiment Analysis * Profiling * Security are some of mobility research concerns that criss-cross agent-oriented systems research and technologies. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 42. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion eLearning Mobility Text Mining Text Mining Text mining is an application field of agents as well as a necessary technology needed by intelligent agents, which may form the basis to dialogue based systems. Knowledge intensive agents are often used in text mining. Agent technologies here include: Blackboard architectures, Description logics, and Ontology. Text mining algorithms may be used in agents forming part of business intelligence system, or agent-based dialogue systems for HCI. Sentiment analysis is currently a hot topic in computational emotions research. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 43. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Emotions in eLearning Emotions in eLearning I Going back to our starting point about the Notion of Agency, and the examples of estate agents, that exact notion makes agent technologies great for delivering services to the end user, i.e. HCI. The flexibility of agent-oriented systems also means that we do not have to develop a full agent or multi-agent system to use agent technology. So whilst it is not necessary to develop a full cognitive agent in an eLearning system interface, emotions and emotion models developed elsewhere can be used. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 44. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Emotions in eLearning Emotions in eLearning II In delivering lectures or labs, a tutor can detect interest and concentration levels in the students. According to this highly heuristic human system, the tutor, often to successful results, adjust the delivery of the lecture to maintain interest in the subject for the majority of the class. eLearning systems in general and Intelligent Tutoring Systems (ITS) in particular lack this much needed ability. Unless ... Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 45. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Emotions in eLearning Emotions in eLearning III We are trying to detect subtle emotional expressions from Facial Expressions to discern the learner emotional state, and hence concentration and interest levels. This focuses on translating facial expressions into emotional model into ”affect” or ”mental” states into an intelligent response by ITS into an adaptive user-centred service delivery. This may be extended later to include Gestures. Then it may be extended further to incorporate work being done at DMU on detecting single emotion, stress, from non-visual devices. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 46. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Conclusion In this talk, we travelled quickly amongst agent technologies bouncing between simplest form of agents as present in swarms to the more complex intelligent variations of cognitive agents. Our focus was on emotions and emotion modelling in particular, which relates strongly to current collaborative project. Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional
  • 47. Agents and Notion of Agency Oscillating Emotions Applied Projects Collaboration with Valencia University Conclusion Questions? Any questions or comments please! And for any follow ups and related publications: Dr. Aladdin Ayesh aayesh@dmu.ac.uk www.aladdin-ayesh.info Dr. Aladdin Ayesh De Montfort University Agent-Oriented Systems: From the Primitive to the Emotional

×