Research

[Work in progress]

Quantum Interaction

Quantum Interaction refers to an emerging field that investigates quantum mechanics (QM) in non-quantum domains such as human language, cognition, information retrieval, artificial intelligence, finance/economics, and social interaction. The Quantum Interaction group can be found here and we also have a wiki.

The QM of Semantic Space

Recently a serendipitous and potentially far reaching connection was made: The formalisation of Quantum Mechanics (QM) shows striking similarities to a class of semantic model emerging from cognitive science. Such models have an impressive track record of replicating aspects of human information processing such as word association norms. This project aims to explore this intriguing connection with the goal of producing a new genre of formal and operational models of human sub-symbolic reasoning related to information processing and retrieval. The general thrust of this research is to provide the groundwork for technology which can genuinely and reliably enhance human awareness in complex information environments.

[The above research is funded by the Australian Research Council Discovery Grant (DP077334)]

Collaborators:

Generalised quantum models of complexity with application to cognitive systems

This project aims to develop a radically new theory for modelling complex systems (e.g., ecosystems, social systems, cognitive systems etc.). Existing theories assume such systems can be separated into interacting parts, often resulting in oversimplified models. The project will provide a significant advance upon current methods by developing novel methods for modelling non-separable systems. The innovation in this project derives from exploiting quantum theory to produce this new genre of modelling tools. The theory will be applied to cognitive systems to underpin frontier information technologies more in tune with humans. More generally the research will open new doors to the modelling and understanding of all complex systems.

[The above research is funded by the Australian Research Council Discovery Grant (DP1094974)]

Collaborators:

Web service discovery

Our view of web service discovery is not only concerned with the retrieval of relevant web-based services but also helping the user discover what it is (s)he needs to know in order to close the agenda at hand. Free text documents in service environments provide an untapped source of information for augmenting the epistemic state of the user and hence their ability come to know what it is they need to know. A quantitative approach to semantic knowledge representation is adopted in the form of cognitively motivated semantic space models. Knowledge of the user’s agenda is promoted by associational inferences computed from the semantic space model. These inferences are abductive in nature and are intended to guide the user from fuzzy search goals to a better understanding of the problem space surrounding the given agenda.

[This project is funded by the Australian Research Council Linkage grant (LP0669924)]

Semantic Profiles

The problem today is that there is little formal and exploratory approach to what information about Interactive Network consumers should be captured and managed. Apart from key personal attributes, there is inconsequential support for wider relationship types and partial profiles. Another key problem is the lack of a formal expression of a consumer’s interests, opinions, and experiences. In essence, apart from checking a consumers contact list, there is little support for consumers to grow their personal and professional connections on Interactive Networks, based on how they interact with the services.

The Technology Challenges for the future include:

  • Mechanisms to support extensible Profile information models to be managed, grouped, and exposed
  • Methods to automatically capture consumers interests, opinions, and experiences based on interactive network behaviour and activities
  • Mechanisms to make connections between Interactive Network consumers that exhibit similar interests, opinions, and experiences

The project focusses the investigation of novel techniques in information retrieval and natural language processing to apply to interactive network activities, such as blogs and discussion groups, to extract succinct interest, opinion, and experience values. We are looking at unique matrix methods and comparative probabilistic techniques to ascertain the best outcomes. The aim is to explore the possibility of linking people who share similar opinions on similar subjects across different communities and across different interactive networks by exploiting the text-based sources of these services.

[This research is partially funded by National ICT Australia (Limited) as part of the SPIN Project]