Research Publications of John Thornton

The Research Page

Papers, Books and Manuscripts

Select By Year:

Select By Subject Area:

Google Scholar

You can also find out about my published research on Google Scholar, which, as well as providing links to the papers and their abstracts, additionally provides the latest citation counts and index values.

2021

Thornton, J. (2021). The Questoning of Intelligence: A phenomenological exploration of what it means to be intelligent. FUBText: Brighton, 511 pp. ISBN 978-1-8384787-0-4

Back Cover: The Questioning of Intelligence is an inquiry by intelligence of intellgence. It is a questioning of the ground on which we understand ourselves and our capacity for intelligent thought and action. Our means of inquiry is the way of phenomenology, the way of entering into the immediacy of being conscious, now. It is from here we can start to investigate the philosophical and scientific inheritance that has formed the collective understanding we currently have of our place in the universe. In questioning this inheritance we are asking after the source from out of which it has emerged, the same source that is manifesting our experience of being alive and conscious now. According to the scientific materialism of our age, this manifestation of experience is no more than an effect of microphysical events occurring in our nervous systems, events that themselves have been determined by an inexorably mechanistic process of physical evolution. It is this materialistic presupposition that stands in the way of recognising the essential form of our natural, innate intelligence. For it is unintelligible to think a system of purely mechanistic calculations could produce the experience of meaning that is the hallmark of human consciousness. Seeing this is not a matter of argument or proof, it is a matter of direct phenomenological insight. It is on the basis of such insight that we look again at the meaning of the findings of contemporary science. For once we put aside this collective materialism, our science reveals an entirely new dimension of significance, where the meaning of our being conscious and intelligent is reflected back in the forms of processes that science has already discovered. From here, perhaps, we can even start to intuit the action of a universal intentionality that expresses itself through these processes, including the processes of our own human consciousness.

Newton, M. A. H., Polash, M. M. A., Pham, D. N., Thornton, J. R., Su, K. & Sattar, A. (2021, In Press). Evaluating Logic Gate Constraints in Local Search for Structured Satisability Problems. Artificial Intelligence Review, Springer, ISSN 0269-2821.

2018

Cowley, B., Thornton, J. R., Main, L. & Sattar, A. (2018). Precision without Precisions: Handling uncertainty with a single predictive model. In: ALIFE 2018, Proceedings of the 2018 Conference on Artificial Life, Tokyo, Japan, MIT Press, pp. 129-136.

2017

Cowley, B., Thornton, J. R., Main, L. & Sattar, A. (2017). Dynamic Thresholds for Self-Organizing Predictive Cells. In: ECAL 2017, Proceedings of the 14th European Conference on Artificial Life, Lyon, France, MIT Press, pp. 114-121.

Cowley, B. & Thornton, J. R. (2017). Feedback Modulated Attention within a Predictive Framework. ACALCI 2017, Artificial Life and Computational Intelligence, 3rd Australasian Conference. Lecture Notes in Computer Science, 10142, 61-73. Springer, ISSN 0302-9743.

Main, L. & Thornton, J. R. (2017). Stable Sparse Encoding for Predictive Processing. In: SSCI 2017, Proceedings of the 2017 IEEE Symposium Series on Computational Intelligence, Hawaii, USA, IEEE, pp. 1-8.

2015

Kneller, A. & Thornton, J. R. (2015). Distal Dendrite Feedback in Hierarchical Temporal Memory. In: IJCNN 2015, Proceedings of the 2015 International Joint Conference on Neural Networks, Killarney, Ireland, IEEE, pp. 385-392.

Main, L. & Thornton, J. R. (2015). A Cortically Inspired Model for Bioacoustics Recognition. ICONIP 2015: 22nd International Conference on Neural Information Processing, Istanbul, Turkey, Lecture Notes in Computer Science, 9492, 348-355, Springer, ISSN 0302-9743.

Thornton, J. R. (2015). The Transcendence of Computational Intelligence. PhD Confirmation Document, 141 pp. Completed under the supervision of John Mandalios, School of Humanities, Griffith University.

Abstract: The nature of the subject matter of this thesis makes it difficult to provide a concise summary of what I intend to write. My aim is rather to discover answers to the questions I am posing through the very process of writing. To begin therefore, I can only state the basic questions that are motivating my work: What is intelligence? What is computation? What is computational intelligence? What could it mean to transcend computational intelligence? Is such transcendence demonstrably possible? Of course I have presentiments of answers to these questions. Firstly, in agreement with Roger Penrose, I would say that any procedure that can be realised as a Turing machine can be considered computational. The question for the thesis is whether we can abstractly consider the functioning of the human brain to be equivalent to that of a (very sophisticated) Turing machine. To answer this I shall be examining the latest hierarchical predictive coding models of the functioning of the human neocortex and asking to what extent such (computational) models can account for the capacities of human intelligence. In particular, I shall be enquiring into our capacity to have a direct knowledge of what it is to be conscious, and asking whether such knowledge provides evidence of our being able to transcend the purely computationally determined actions of our neural systems. The confirmation document itself comprises an introductory chapter which provides an outline of the final thesis. At the moment I envisage a three-part manuscript. Part One is concerned with the problem of gaining explicit access to a state of pure, pre-reflective consciousness and thereby discovering what it means to have a direct knowledge of consciousness. In Part Two I plan to examine the computational foundations of intelligence in the functioning of the human neocortex, both from a neuroscientific and phenomenological perspective. Part Three is then intended to draw together Parts One and Two in a consideration of the possible transcendence of computational intelligence.
Following on from the introduction, Chapters 2-5 are all intended for inclusion in the final thesis. They cover Part One described above, with Chapter 2 providing a method of access, and Chapters 3-5 examining the work of Descartes, Schopenhauer and Husserl in relation to the material presented in Chapter 2. Finally, the appendices present two conference papers and one workshop paper that I have written and presented during my candidature. These papers cover various aspects of the questions I intend to address in Parts Two and Three, including the relevance of Heidegger to contemporary cognitive neuroscience (Appendix A), whether the activity of the human brain is causally closed under laws that determine the local low-level functioning of neural populations (Appendix B) and how contemporary models of neocortical functioning can be mapped onto Husserl’s phenomenological account of temporal consciousness (Appendix C).

2014

Cai, S., Luo, C., Thornton, J. R. & Su, K. (2014). Tailoring Local Search for Partial MaxSAT. In: AAAI 2014, Proceedings of the 28th Conference on Artificial Intelligence, Quebec, Canada, MIT Press, pp. 2623-2629.

Cowley, B., Kneller, A. & Thornton, J. R. (2014). Cortically-Inspired Overcomplete Feature Learning for Colour Images. PRICAI 2014: Trends in Artificial Intelligence, 13th Pacific Rim International Conference on Artificial Intelligence. Lecture Notes in Computer Science, 8862, 720-732, Springer, ISSN 0302-9743.

Thornton, J. R. (2014). Hierarchical Temporal Intentionality. Abstract in: ASSC 18: Handbook of the 18th Conference of the Association for the Scientific Study of Consciousness, Brisbane, Australia, University of Queensland, pp. 41-42.

2013

Thornton, J. R. & Srbic, A. (2013). Spatial Pooling for Greyscale Images. International Journal of Machine Learning and Cybernetics. 4(3), 207-216, Springer, ISSN 1868-8071.

2012

Ramachandran, R., Wang, K., Wang, J. & Thornton, J. R. (2012). Probabilistic Reasoning in DL-Lite. PRICAI 2012: 12th Pacific Rim International Conference on Artificial Intelligence, Lecture Notes in Computer Science. 7458, 480-491, Springer, ISSN 0302-9743.

Thornton, J. R., Main L. & Srbic, A. (2012). Fixed Frame Temporal Pooling. AI 2012: 25th Australasian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 7691, 707-718, Springer, ISSN 0302-9743.

Thornton, J. R. (2012). The Phenomenological Negation of the Causal Closure of the Physical. Abstract in: AAP 2012: Proceedings of the 2012 Conference of the Australasian Association for Philosophy, Wollongong, Australia, Wollongong University, p. 22.

Thornton, J. R. (2012). The Phenomenological Negation of Objective Physicalism. Unpublished philosophical paper. 35 pp. Completed under the supervision of Deborah Brown at the School of History, Philosophy, Religion and Classics, University of Queensland, Brisbane.

Abstract: An outstanding task for contemporary philosophy of mind and cognitive science is to evaluate the significance and relevance of the phenomenological tradition for the study of mind and consciousness. At the moment, mainstream philosophy of mind stands in relative ignorance of the original work of the key phenomenologists, content either to dismiss phenomenology as muddled and self-contradictory, or to understand it in terms of the interpretations of a few famous contemporaries. Insofar as phenomenology is accepted into the analytic debate, it is seen as playing a supporting role, generating the empirical data of accurately observed experience that then acts as a test of adequacy for analytic theory. The revolutionary aspects of phenomenology, the transcendentalism of Husserl, the destruction of traditional ontology by Heidegger, are placed firmly on the other side of the continental divide.
In this paper, I intend to show that authentic phenomenology stands in a fundamental opposition to the assumptions and practices of contemporary analytic philosophy of mind. Phenomenology is not simply a useful tool for generating observations of subjective experience, it is way of engaging in philosophical enquiry that moves in an opposite direction to analytic thought. Unless this is seen clearly, the pronouncements of the phenomenologists will be misinterpreted and inappropriately rejected. To understand phenomenology, one has to enter into phenomenological enquiry ‘feet first.’ It is not a matter of a theoretical understanding. One of the aims of the phenomenological method is to uncover the pre-theoretical ground of theoretical understanding. Arriving at such a pre-theoretical ground means suspending one’s theoretical perspective. And what it means to suspend one’s theoretical perspective is itself a phenomenological question.
The point is that a positive theoretical account of the phenomenological method is not going to communicate what it means to uncover the pre-theoretical ground of experience. One has to proceed on the basis of experience and understand on the basis of negation. These principles can only be claried through actual demonstration. This is analogous to what Mary learned on seeing colour for the rst time. My plan is to illustrate the phenomenological method through a phenomenological analysis of the general framework within which contemporary philosophy of mind operates. I shall term this general framework objective physicalism. What exactly is meant by objective physicalism will become clearer as we proceed.

2011

Sotomayor, M., Wang, K., Shen, Y. & Thornton, J. R. (2011). Probabilistic Multi-context Systems. JIST 2011: Joint International Semantic Technology Conference. Lecture Notes in Computer Science, 7185, 366-375, Springer, ISSN 0302-9743.

Thornton, J. R., Srbic, A., Main, L. & Chitsaz, M. (2011). Augmented Spatial Pooling. AI 2011: 24th Australasian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science. 7106, 261-270, Springer, ISSN 0302-9743.

Thornton, J. R. (2011). The Consciousness Test. Unpublished philosophical paper. 25 pp. Completed under the supervision of Bruin Christensen and David Chalmers at the School of Philosophy, Australian National University, Canberra.

2010

Thornton, J. R. & Christensen, C. B. (2010). An Essential Difference: Wheeler and Heidegger on the relationship between science and philosophy. Presented at: Reconstructing the Cognitive World: A workshop with Michael Wheeler, Goethe University Frankfurt am Main, February 3-4, 2010.

Introduction: Michael Wheeler, in his book Reconstructing the Cognitive World, analyses the development of embedded-embodied cognitive science in the light of underlying philosophical differences about the constitution of human agency. On one side he sees orthodox computational cognitive science as holding to Cartesian conceptions of an abstract, disembodied reason deliberating over de-contextualised representations of the world. On the other side, he sees modern-day embodied-embedded cognitive scientists going beyond such Cartesianism to embrace concepts of human agency more in keeping with Heidegger’s account of Dasein in Being and Time. By bringing to light and criticising the Cartesian assumptions of the computationalists and by pointing out and clarifying the connections between embodied-embedded cognitive science and Heidegger’s philosophy, Wheeler aims to lay the “foundations of a genuinely non-Cartesian cognitive science.”
Along the way, Wheeler argues that Heidegger is a scientific realist who holds that modern science provides genuinely objective epistemic access to independently real entities. He takes this to show that Heidegger would not object to the incorporation of his account of Dasein into the broad framework of contemporary cognitive-scientific explanation. According to Wheeler, such explanation is:

. . . a species of empirical explanation in which the ultimate goal is to map out the subagential elements (e.g., the neural states and mechanisms, or the functionally identified psychological subsystems) whose organization, operation, and interaction make it intelligible to us how it is that unmysterious causal processes (such as those realized in brains) can give rise to psychological phenomena that are genuinely constitutive of agency and cognition.

Within this framework, cognitive scientists necessarily make assumptions about what the relevant psychological phenomena are and how they are constitutive of agency and cognition. It is in the formulation of these assumptions that Wheeler sees the hidden hand of Descartes and the potential for a new Heideggerian reconstruction of the area. For Wheeler, the major philosophical task is to clarify these assumptions, bringing into question existing orthodox views and providing a philosophical underpinning for newer embodied-embedded conceptions of agency. Evidently, this task commits Wheeler to endorsing, at least in general terms, the same framework and thus the same assumptions that underlie what he defines as the goal of cognitive-scientific explanation. Wheeler supposes that this presents no problems for the claim that much from Heidegger’s philosophy can be incorporated into cognitive science. To support this, Wheeler adduces a number of passages from Being and Time in which Heidegger appears to endorse Wheeler’s views on philosophy’s role in identifying and clarifying the constitutive assumptions of individual sciences. And so Wheeler concludes that how he conceives such philosophical clarification is basically how Heidegger conceives it.
In this paper we shall argue that this is not correct and that the key reason for this turns on the issue of naturalism. According to Wheeler any serious attempt to achieve the defining goal of cognitive science must accept a commitment to naturalism. Yet the commitment Wheeler appears to have to naturalism (as we shall see, it is not quite clear what this commitment is) and certainly the unquestioned character of it, separates him from Heidegger. In consequence, the actual affinities between Heidegger and Wheeler are limited because Wheeler accepts as given something which, at least on any strong reading of it, Heidegger is concerned to dislodge. This difference entails significant divergence in how each sees the relation of philosophy to science.

2009

Thornton, J. R., Faichney, J., Blumenstein, M., Nguyen, V. & Hine, T. (2009). Offline Cursive Character Recognition: A state-of-the-art comparison. In: IGS 2009: Proceedings of the 14th Conference of the International Graphonomics Society, Dijon, France, pp. 148-151.

2008

Pham, D. N., Thornton, J. R., Gretton, C. & Sattar, A. (2008). Combining Adaptive and Dynamic Local Search for Satisfiability. Journal on Satisfiability, Boolean Modeling and Computation. 4, 149-172, The SAT Association, ISSN 1574-0617.

Pham, D. N., Thornton, J. R. & Sattar, A. (2008). Efficiently Exploiting Dependencies in Local Search for SAT. In AAAI 2008: Proceedings of the 23rd Conference on Artificial Intelligence, MIT Press, pp. 1476-1478.

Pham, D. N., Thornton, J. R. & Sattar, A. (2008). Modelling and Solving Temporal Reasoning as Propositional Satisfiability. Artificial Intelligence. 172(15), 1752-1782, ISSN 0004-3702.

Abstract: Representing and reasoning about time dependent information is a key research issue in many areas of computer science and artificial intelligence. One of the best known and widely used formalisms for representing interval-based qualitative temporal information is Allen’s interval algebra (IA). The fundamental reasoning task in IA is to find a scenario that is consistent with the given information. This problem is in general NP-complete.
In this paper, we investigate how an interval-based representation, or IA network, can be encoded into a propositional formula of Boolean variables and/or predicates in decidable theories. Our task is to discover whether satisfying such a formula can be more efficient than finding a consistent scenario for the original problem. There are two basic approaches to modelling an IA network: one represents the relations between intervals as variables and the other represents the end-points of each interval as variables. By combining these two approaches with three different Boolean satisfiability (SAT) encoding schemes, we produced six encoding schemes for converting IA to SAT. In addition, we also showed how IA networks can be formulated into satisfiability modulo theories (SMT) formulae based on the quantifier-free integer difference logic (QF-IDL). These encodings were empirically studied using randomly generated IA problems of sizes ranging from 20 to 100 nodes. A general conclusion we draw from these experimental results is that encoding IA into SAT produces better results than existing approaches. More specifically, we show that the new point-based 1-D support SAT encoding of IA produces consistently better results than the other alternatives considered. In comparison with the six different SAT encodings, the SMT encoding came fourth after the point-based and interval-based 1-D support schemes and the point-based direct scheme. Further, we observe that the phase transition region maps directly from the IA encoding to each SAT or SMT encoding, but, surprisingly, the location of the hard region varies according to the encoding scheme. Our results also show a fixed performance ranking order over the various encoding schemes.

Thornton, J. R. & Pham, D. N. (2008). Using Cost Distributions to Guide Weight Decay in Local Search for SAT. PRICAI 2008: 10th Pacific Rim International Conference on Artificial Intelligence. Lecture Notes in Computer Science, 5351, 405-416, Springer, ISSN 0302-9743.

Thornton, J. R., Faichney, J., Blumenstein, M. & Hine, T. (2008). Character Recognition using Hierarchical Vector Quantization and Temporal Pooling. AI 2008: 21st Australasian Joint Conference on Artificial Intelligence. Lecture Notes in Computer Science, 5360, 562-572, Springer, ISSN 0302-9743.

2007

Ishtaiwi, A., Thornton, J. R. & Sattar, A. (2007). Weight Redistribution for Unweighted MAX-SAT. AI 2007: Advances in Artificial Intelligence, 20th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 4830, 687-693, Springer, ISSN 0302-9743.

Orgun, M. A. & Thornton, J. (Eds.) (2007). AI 2007: Advances in Artificial Intelligence. Lecture Notes in Artificial Intelligence, 4830, Heidelberg: Springer, ISBN 978-3-540-76926-2.

Preface: This volume contains the papers presented at AI 2007: The 20th Australian Joint Conference on Artificial Intelligence held during December 2–6, 2007 on the Gold Coast, Queensland, Australia. AI 2007 attracted 194 submissions (full papers) from 34 countries. The review process was held in two stages. In the first stage, the submissions were assessed for their relevance and readability by the Senior Program Committee members. Those submissions that passed the first stage were then reviewed by at least three Program Committee members and independent reviewers. After extensive discussions, the Committee decided to accept 60 regular papers (acceptance rate of 31%) and 44 short papers (acceptance rate of 22.7%). Two regular papers and four short papers were subsequently withdrawn and are not included in the proceedings. AI 2007 featured invited talks from four internationally distinguished researchers, namely, Patrick Doherty, Norman Foo, Richard Hartley and Robert Hecht-Nielsen. They shared their insights and work with us and their contributions to AI 2007 were greatly appreciated. AI 2007 also featured workshops on integrating AI and data-mining, semantic biomedicine and ontology. The short papers were presented in an interactive poster session and contributed to a stimulating conference.
We would like to thank the Conference Co-chairs, Abdul Sattar and Vladimir Estivill-Castro of Griffith University for their guidance, and the local Organizing Co-chairs Michael Blumenstein and Guido Governatori for making sure that the conference ran smoothly. Special thanks go to Natalie Dunstan and Vicky Wheeler for supporting the Committees so effectively. We also would like to thank the following organizations for their generous sponsorship of AI 2007: Griffith University, National ICT Australia, the University of Queensland, Queensland University of Technology, Bond University, the Australian Computer Society and the Gold Coast City Council.

Pham, D. N., Thornton, J. R., Gretton, C. & Sattar, A. (2007). Advances in Local Search for Satisfiability. AI 2007: 20th Australian Joint Conference on Artificial Intelligence. Lecture Notes in Computer Science, 4830, 213-222, Springer, ISSN 0302-9743.

Pham, D. N., Thornton, J. R. & Sattar, A. (2007). Building Structure into Local Search for SAT. In: IJCAI 2007, Proceedings of the 20th International Joint Conference on Artificial Intelligence, pp. 2359-2364. Winner of Distinguished Paper Award.

Thornton, J. (2007). The Impersonal Knowledge of Conscious Experience: A philosophical investigation. Unpublished philosophical manuscript. 81 pp.

Preface: My motivation in writing this manuscript is to fundamentally challenge the prevailing wisdom of the philosophy of mind and artificial intelligence communities that conscious experience is created or caused by the operation of physical neurons in a physical brain. Because this idea remains unchallenged, it is now widely treated as an obvious truth, a fact beyond dispute. And so it is disseminated into the larger human community, through the media, and through scientific education, without further reflection.
As I shall demonstrate, this naturalistic scientific conception of consciousness is little more than an assertion of faith or belief, which remains unsupported by factual evidence. To see how this has come about, and how the reality of consciousness remains beyond the grasp of contemporary scientific and philosophical thought, requires that we each, individually, investigate this issue in our own conscious experience.
My aim is therefore not to give a detailed criticism of the particular arguments and positions of contemporary thinkers but to present a methodology for the investigation of the mind. The basic thrust of this methodology is the attainment of an impersonal first-person observational standpoint, and to show how such a standpoint can act as the foundation for an authentic philosophy of mind. The idea is to proceed by practical demonstration and not by conceptual argumentation. As such, this work flies in the face of current practice, and there is much in it that will appear provocative and even nonsensical to someone schooled in the existing literature.
Therefore I have to ask my prospective readers to suspend their judgement, as far as is possible, until they have absorbed what has been described in the following pages. For this is a practical work, and asks you to test what is proposed by actual observation and not according to the thoughts and opinions of others.
I should make it clear that this approach to the investigation of consciousness has not been plucked out of thin air. It owes much to the phenomenology of Edmund Husserl, although it goes somewhat further in identifying thought as the agency of the personal viewpoint. Behind that lies the philosophy of Arthur Schopenhauer and my quite separate experience of the work of various twentieth century spiritual teachers. Of these the most profound inspiration has been the life and work of Barry Long (1926-2003). In particular, it is Barry Long’s recognition of the fundamental distinction between thought and observation that lies at the core of the current work. Not that you will find a spiritual teaching in these pages. This is a work of philosophy and does not assume the existence of any spiritual realm. Quite the reverse. What is asked for here is only that we remain within the empirical realm of our direct, first-person experience.

Thornton, J. (2007). The Foundations of Computing and the Information Technology Age: A historical, sociological and philosophical enquiry. Pearson Education Australia, 286 pp. ISBN 978-0-7339-8848-6

Back Cover: The Foundations of Computing and the Information Technology Age is a book both for undergraduate computing students and for anyone seeking a deeper understanding of technology in the modern world. Dispensing with simplistic explanations, the book first considers the evolution of the computer from the origins of number to the development of the microprocessor. Along the way we meet the early pioneers of mechanical calculation, including Pascal and Leibniz, the groundbreaking work of Charles Babbage and his Difference Engines and the drama of the wartime code-breakers at Bletchley Park.
But this is not just a historical text. It provides an introduction to the theory of computation, showing how Alan Turing’s concept of a universal Turing machine helped form the foundations of modern computer science. Theory then becomes practice as the book explores the von Neumann architecture and shows how simple switching circuits can be used to construct a general purpose computer.
The basic theme running throughout this discussion is that the foundations of computing and the information technology age lie in the scientific turn of mind taken by our entire civilisation. From this perspective, the book traces how information technology has been used to restructure the economic and social life of the developed world and enquires into the ultimate direction and purpose of this process of globalisation. The reader is then drawn to consider how our technical, materialistic understanding has ignored the underlying reality from which all technology emerges: human consciousness.
Finally, the book argues that this inability to acknowledge the central reality of consciousness has caused modern civilisation to enter into an unbalanced pattern of development, where we increasingly understand ourselves as biological machines that must be adapted to the latest technology, rather than as the creative intelligence that technology was originally supposed to serve.

2006

Ishtaiwi, A., Thornton, J. R., Anbulagan, Sattar, A. & Pham, D. N. (2006). Adaptive Clause Weight Redistribution. CP 2006: 12th International Conference on the Principles and Practice of Constraint Programming. Lecture Notes in Computer Science, 4204, 229-243, Springer, ISSN 0302-9743.

Pham, D. N., Thornton, J. R. & Sattar, A. (2006). Towards an Efficient SAT Encoding for Temporal Reasoning. CP 2006: 12th International Conference on the Principles and Practice of Constraint Programming. Lecture Notes in Computer Science, 4204, 421-436, Springer, ISSN 0302-9743.

Thornton, J. R., Gustafsson, T., Blumenstein, M. & Hine, T. (2006). Robust Character Recognition using a Hierarchical Bayesian Network. AI 2006: 19th Australian Joint Conference on Artificial Intelligence. Lecture Notes in Computer Science, 4304, 1259-1264, Springer, ISSN 0302-9743.

2005

Bain, S., Thornton, J. R. & Sattar, A. (2005). Evolving Variable-Ordering Heuristics for Constrained Optimisation. CP 2005: 11th International Conference on Contraint Programming. Lecture Notes in Computer Science, 3709, 732-736, Springer, ISSN 0302-9743.

Bain, S., Thornton, J. R. & Sattar, A. (2005). A Comparison of Evolutionary Methods for the Discovery of Local Search Heuristics. AI 2005: 18th Australian Joint Conference on Artificial Intelligence. Lecture Notes in Computer Science, 3809, 1068-1074, Springer, ISSN 0302-9743.

Ferreira Jr., V. & Thornton, J. R. (2005). Tie Breaking in Clause Weighting Local Search for SAT. AI 2005: 18th Australian Joint Conference on Artificial Intelligence. Lecture Notes in Computer Science, 3809, 70-81, Springer, ISSN 0302-9743.

Ishtaiwi, A., Thornton, J. R., Sattar, A. & Pham, D. N. (2005). Neighbourhood Clause Weight Redistribution in Local Search for SAT. CP 2005: 11th International Conference on the Principles and Practice of Constraint Programming. Lecture Notes in Computer Science, 3709, 772-776, Springer, ISSN 0302-9743.

Pham, D. N., Thornton, J. R., Sattar, A. & Ishtaiwi, A. (2005). SAT-based versus CSP-based Constraint Weighting for Satisfiability. In AAAI 2005: Proceedings of the 20th National Conference on Artificial Intelligence, MIT Press, pp. 455-460.

Pham, D. N., Thornton, J. R. & Sattar, A. (2005). Modelling and Solving Temporal Reasoning as Satisfiability. Proceedings of the 4th International Workshop on Modelling and Reformulating Constraint Satisfaction Problems, Sitges, Spain. pp. 117-131.

Thornton, J. R. (2005). Clause Weighting Local Search for SAT. Journal of Automated Reasoning. 35(1-3), 97-142, Springer, ISSN 0168-7433.

Abstract: This paper investigates the necessary features of an effective clause weighting local search algorithm for propositional satisfiability testing. Using the recent history of clause weighting as evidence, we suggest that the best current algorithms have each discovered the same basic framework, i.e. to increase weights on false clauses in local minima and then to periodically normalize these weights using a decay mechanism.
Within this framework, we identify two basic classes of algorithm according to whether clause weight updates are performed additively or multiplicatively. Using one of the best recently developed multiplicative algorithms (SAPS) and our own pure additive weighting scheme (PAWS), an experimental study was constructed to isolate the effects of multiplicative in comparison to additive weighting, while controlling other key features of the two approaches, namely the use of pure versus flat random moves, deterministic versus probabilistic weight smoothing and multiple versus single inclusion of literals in the local search neighbourhood. In addition, we examined the effects of adding a threshold feature to multiplicative
weighting that makes it indifferent to similar cost moves.
As a result of this investigation, we show that additive weighting can outperform multiplicative weighting on a range of difficult problems, while requiring considerably less effort in terms of parameter tuning. Our examination of the differences between SAPS and PAWS suggests that additive weighting does benefit from the random flat move and deterministic smoothing heuristics, whereas multiplicative weighting would benefit from a deterministic/probabilistic smoothing switch parameter that is set according to the problem instance. We further show that adding a threshold to multiplicative weighting produces a general deterioration in performance, contradicting our earlier conjecture that additive weighting has better performance due to having a larger selection of possible moves. This leads us to explain differences in performance as being mainly caused by the greater emphasis of additive weighting on penalizing clauses with relatively less weight.

2004

Bain, S., Thornton, J. R. & Sattar, A. (2004). Methods of Automatic Algorithm Generation. PRICAI 2004: 8th Pacific Rim International Conference on Artificial Intelligence. Lecture Notes in Computer Science, 3157, 144-153, Springer, ISSN 0302-9743.

Bain, S., Thornton, J. R. & Sattar, A. (2004). Evolving Algorithms for Constraint Satisfaction. In: CEC 2004: Proceedings of the 2004 Congress on Evolutionary Computation, IEEE, pp. 265-272.

Beaumont, M., Thornton, J. R., Sattar, A. & Maher, M. (2004). Solving Over-constrained Temporal Reasoning Problems using Local Search. PRICAI 2004: 8th Pacific Rim International Conference on Artificial Intelligence, Lecture Notes in Computer Science, 3157, 134-143, Springer, ISSN 0302-9743..

Ferreira Jr., V. & Thornton, J. R. (2004). Longer-Term Memory in Clause Weighting Local Search for SAT. In AI 2004: 17th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 3339, 730-741, Springer, ISSN 0302-9743.

Stantic, B., Khanna, S. & Thornton, J. R. (2004). An Efficient Method for Indexing Now-relative Bitemporal Data. In ADC 2004: Proceedings of the 15th Australasian Database Conference, Australian Computer Society, pp. 113-122.

Thornton, J. R., Pham, D. N., Bain, S. & Ferreira Jr., V. (2004). Additive versus Multiplicative Clause Weighting for SAT. In AAAI 2004: Proceedings of the 19th National Conference on Artificial Intelligence, The MIT Press, pp. 191-196.

Thornton, J. R., Beaumont, M., Sattar, A. & Maher, M. (2004). A Local Search Approach to Modelling and Solving Interval Algebra Problems. Journal of Logic and Computation, 14(1), 93-112, Oxford University Press, ISSN 0955-792X.

Zhou, L., Thornton, J. R. & Sattar, A. (2004). Dynamic Agent-Ordering and Nogood-Repairing in Distributed Constraint Satisfaction Problems. In FLAIRS 2004: Proceedings of the 17th International Florida Artificial Intelligence Research Society Conference, AAAI Press, pp. 20-25.

2003

Anbulagan, Thornton, J. R., & Sattar, A. (2003). Dynamic Variable Filtering for Hard Random 3-SAT Problems. AI 2003: 16th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2903, 100-111, Springer, ISSN 0302-9743.

Abstract: The DPL (Davis-Putnam-Logemann-Loveland) procedure is one of the most effective methods for solving SAT problems. It is well known that its efficiency depends on the choice of the branching rule. One of the main improvements of this decision procedure has been the development of better branching variable selection through the use of unit propagation look-ahead (UPLA) heuristics (e.g., [13]). UPLA heuristics perform one of the following actions during two propagations of a free variable at each search tree node: detecting a contradiction earlier, simplifying the formula, or weighing the branching variable candidates. UPLA heuristics can be considered as polynomial time reasoning techniques. In this paper, we propose a new branching rule which does more
reasoning to narrow the search tree of hard random 3-SAT problems. We term this reasoning technique the dynamic variable filtering heuristic. In our empirical study we develop four different variants of the DPL procedure: two (ssc34 and ssc355) based on this new heuristic and another two (Satz215-0 and Satz215sT ) based on static variable filtering heuristics. ssc355 additionally integrates the Neighborhood Variable Ordering (NVO) heuristic into ssc34. We then compare the best known versions of the state-of-the-art Satz DPL procedure (Satz215), with each of our four procedures. Our results suggest that improved branching rules can further enhance the performance of DPL procedures. On hard random 3-SAT problems, our best procedure (ssc355) outperforms Satz215 with an order of magnitude in terms of the number of branching nodes in the search tree. While the run-times for dynamic variable filtering are still uncompetitive with Satz215, we have yet to explore the benefits that can be gained from avoiding redundant propagations and we still can improve the performance of the NVO heuristic. A further interesting property of dynamic filtering is that all backtracking can be eliminated during the DPL unit rule process. This property can also be explored in our future work for improving DPL procedure efficiency.

Leonard, J., Treffner, P., & Thornton, J. R. (2003). Tau Guidance for Mobile Soccer Robots. Studies in Perception and Action VII, 169-172, Psychology Press, ISBN 978-0805848052.

Pullan, W., Zhao, L., & Thornton, J. R. (2003). Estimating Problem Metrics for SAT Clause Weighting Local Search. AI 2003: 16th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2903, 137-149, Springer, ISSN 0302-9743.

Stantic, B., Thornton, J. R., & Sattar, A. (2003). A Novel Approach to Model NOW in Temporal Databases. In TIME-ICTL 2003: Proceedings of the 10th International Symposium on Temporal Representation and Reasoning / 4th International Conference on Temporal Logic, IEEE, pp. 174-181.

Zhou, L., Thornton, J. R., & Sattar, A. (2003). Dynamic Agent Ordering in Distributed Constraint Satisfaction Problems. AI 2003: 16th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2903, 427-439, Springer, ISSN 0302-9743.

2002

Ferreira Jr., V., Thornton, J. R. & Leonard, J. (2002). A Subsumption Architecture for Robotic Soccer. In: Proceedings of the 2002 FIRA Robot World Congress, KAIST (Korea Advanced Institute of Science and Technology), pp. 648-654.

Kravchuk, O., Pullan, W., Thornton, J. R. & Sattar, A. (2002). An Investigation of Variable Relationships in 3-SAT Problems. AI 2002: 15th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2557, 579-590, Springer, ISSN 0302-9743.

Thornton, J. R., Bain, S., Sattar, A. & Pham, D. (2002). A Two Level Local Search for MAX-SAT Problems with Hard and Soft Constraints. AI 2002: 15th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2557, 603-614, Springer, ISSN 0302-9743.

Thornton, J. R., Beaumont, M., Sattar, A. & Maher, M. (2002). Applying Local Search to Temporal Reasoning. In: TIME 2002: 9th International Symposium on Temporal Reasoning and Representation, IEEE, pp. 94-99, ISSN 1530-1311.

Thornton, J. R., Leonard, J., Wiseby, R. & Lee, Y. (2002). Shape Recognition and Enhanced Control Systems for Robot Soccer. In: Proceedings of the 2002 FIRA Robot World Congress, KAIST (Korea Advanced Institute of Science and Technology), pp. 670-675.

Thornton, J. R., Pullan, W. & Terry, J. (2002). Towards Fewer Parameters for Clause Weighting SAT Algorithms. AI 2002: 15th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2557, 603-614, Springer, ISSN 0302-9743.

2001

Beaumont, M., Sattar, A., Maher, M., & Thornton, J. R. (2001). Solving Overconstrained Temporal Reasoning Problems. AI 2001: 14th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 2556, 603-614, ISSN 0302-9743.

2000

Nagarajan, S., Goodwin, S., Sattar, A, & Thornton, J. R. (2000). On Dual Encodings for Non-Binary Constraint Satisfaction Problems. CP 2000: 6th International Conference on the Principles and Practice of Constraint Programming. Lecture Notes in Computer Science, 1894, 531-536, Spinger, ISSN 0302-9743.

Thornton, J. R. (2000). Contraint Weighting Local Search for Constraint Satisfaction. PhD Thesis, School of Computing and Information Technology, Griffith University, Brisbane, Australia, pp. 160.

Abstract: One of the challenges for the constraint satisfaction community has been to develop an automated approach to solving Constraint Satisfaction Problems (CSPs) rather than creating specific algorithms for specific problems. Much of this work has concentrated on the development and improvement of general purpose backtracking techniques. However, the success of relatively simple local search techniques on larger satisfiability problems [Selman et al. 1992] and CSPs such as the n-queens [Minton et al. 1992] has caused interest in applying local search to constraint satisfaction. In this thesis we look at the usefulness of constraint weighting as a local search technique for constraint satisfaction. The work is based on the clause weighting ideas of Selman and Kautz [1993] and Morris [1993] and applies, evaluates and extends these ideas from the satisfiability domain to the more general domain of CSPs. Specifically, the contributions of the thesis are:

The introduction of a local search taxonomy. We examine the various better known local search techniques and recognise four basic strategies: restart, randomness, memory and weighting.
The extension of the CSP modelling framework. In order to represent and efficiently solve more realistic problems we extend the CSP modelling framework to include array-based domains and array-based domain use constraints.
The empirical evaluation of constraint weighting. We compare the performance of three constraint weighting strategies on a range of CSP and satisfiability problems and with several other local search techniques. We find that no one technique dominates in all problem domains.
The characterisation of constraint weighting performance. Based on our empirical study we identify the weighting behaviours and problem features that favour constraint weighting. We conclude weighting does better on structured problems where the algorithm can recognise a harder sub-group of constraints.
The extension of constraint weighting. We introduce an efficient arc weighting algorithm that additionally weights connections between constraints that are simultaneously violated at a local minimum. This algorithm is empirically shown to outperform standard constraint weighting on a range of CSPs and within a general constraint solving system. Also we look at combining constraint weighting with other local search heuristics and find that these hybrid techniques can do well on problems where the parent algorithms are evenly matched.
The application of constraint weighting to over constrained domains. Our empirical work suggests constraint weighting does well for problems with distinctions between constraint groups. This led us to investigate solving real-world over constrained problems with hard and soft constraint groups and to introduce two dynamic constraint weighting heuristics that maintain a distinction between hard and soft constraint groups while still adding weights to violated constraints in a local minimum. In an empirical study, the dynamic schemes are shown to outperform other fixed weighting and non-weighting systems on a range of real world problems. In addition, the performance of weighting is shown to degrade less severely when soft constraints are added to the system, suggesting constraint weighting is especially applicable to realistic, hard and soft constraint problems.

1999

Thornton, J. R., & Sattar, A. (1999). On the Behaviour and Application of Constraint Weighting. CP 1999: 5th International Conference on the Principles and Practice of Constraint Programming, Lecture Notes in Computer Science, 1713, 446-460, Springer, ISSN 0302-9743.

1998

Thornton, J. R., & Sattar, A. (1998). Using Arc Weights to Improve Iterative Repair. In AAAI 1998: 15th National Conference on Artificial Intelligence, MIT Press, pp. 367-372.

Thornton, J. R., & Sattar, A. (1998). Dynamic Constraint Weighting for Over Constrained Problems. PRICAI 1998: 5th Pacific Rim International Conference on Artificial Intelligence, Lecture Notes in Computer Science, 1531, 377-388, Springer, ISSN 0302-9743.

1997

Thornton, J. R., & Sattar, A. (1997). Applied Partial Constraint Satisfaction Using Weighted Iterative Repair. AI 1997: 10th Australian Joint Conference on Artificial Intelligence, Lecture Notes in Computer Science, 1342. Springer, ISSN 0302-9743.

Thornton, J. R., & Sattar, A. (1997). Nurse Rostering and Integer Programming Revisited. In: ICCIMA 1997: Proceedings of the International Conference on Computational Intelligence and Multimedia Applications, Griffith University, pp. 49-58.

1996

Thornton, J. R., & Sattar, A. (1996). An Integer Programming-Based Nurse Rostering System. In ASIAN 1996: 2nd Asian Computing Science Conference, Lecture Notes in Computer Science, 1179, 357-358, Springer, ISSN 0302-9743.

1995

Thornton, J. R. (1995). An Enhanced Cyclic Descent Algorithm for Nurse Rostering. Honours Thesis, Faculty of Engineering and Applied Science, Griffith University, Gold Coast, Australia, 145 pp.

Abstract: The study introduces an enhanced cyclic descent algorithm for nurse rostering. The algorithm is compared to four other rostering algorithms and to manually generated roster solutions obtained from the Gold Coast Hospital. Three criteria are developed with which the roster generation methods are assessed: these are roster schedule quality, roster shift allocation quality and execution time. A statistical analysis shows that the enhanced cyclic descent algorithm has the best overall performance. An integer linear programming algorithm and an enhanced simulated annealing algorithm are also shown to perform well with smaller problems.