Crime investigation is a complex task involving vast amounts of information and requiring many di... more Crime investigation is a complex task involving vast amounts of information and requiring many different types of expert knowledge. Crime investigators would therefore benefit from the use of decision support systems to help manage this in-formation and to provide knowledge to help solve the more complex problems. Current research efforts in this area have focussed on the information management side
Probabilistic abduction extends conventional symbolic abductive reasoning with Bayesian inference... more Probabilistic abduction extends conventional symbolic abductive reasoning with Bayesian inference methods. This allows for the uncertainty underlying implications to be expressed with probabilities as well as assumptions, thus complementing the symbolic approach in situations where the use of a complete list of assumptions underlying inferences is not practical. However, probabilistic abduction has been of little use in first principle-based applications, such as abductive diagnosis, largely because no methods are available to automate the construction of probabilistic models, such as Bayesian networks (BNs). This paper addresses this issue by proposing a compositional modelling method for BNs.
Despite increasing interest in the development of intelligent techniques to aid in the prevention... more Despite increasing interest in the development of intelligent techniques to aid in the prevention and detection of crime, an important issue that has not yet been addressed by existing work is the use of knowledge based systems (KBS) to aid in the synthesis and analysis of hypothetical scenarios in major crime investigation. The main limitation of conventional KBS approaches is their lack of robustness to deal with the substantial variability of crime scenarios. This paper introduces a method to apply model based reasoning techniques to this problem. In particular, the existing compositional Bayesian modelling approach is extended and adapted to create hypothetical crime scenarios. Also, methods developed in the area of model-based diagnosis are used to support the analysis of synthesised crime scenarios.
... Cambridge University Press, 1969. 23. J. Venn. The Logic of Chance: An Essay on the Foundatio... more ... Cambridge University Press, 1969. 23. J. Venn. The Logic of Chance: An Essay on the Foundations and Province of the Theory of Probability. Macmillan, 1866. 24. J. von Plato. definetti's earliest works on the foundations of probability. Erkenntnis, 31:263-282, 1989. ...
An important cause of miscarriages of justice is the failure of crime investigators and lawyers t... more An important cause of miscarriages of justice is the failure of crime investigators and lawyers to consider important plausible explanation for the available evidence. Recent research has explored the development of decision support systems that (i) assist human crime investigators by proposing plausible crime scenarios explaining given evidence, and (ii) provide the means to analyse such scenarios. While such approaches can generate useful explanations, they are inevitably restricted by the limitations of formal abductive inference mechanisms. Building on work presented previously at this venue, this paper characterises an important class of scenarios, containing "alternative suspects" or "hidden objects", which cannot be synthesised robustly using conventional abductive inference mechanisms. The work is then extended further by proposing a novel inference mechanism that enables the generation of such scenarios.
Norms are a valuable mechanism for establishing coherent cooperative behaviour in decentralised s... more Norms are a valuable mechanism for establishing coherent cooperative behaviour in decentralised systems in which no central authority exists. In this context, Axelrod's seminal model of norm establishment in populations of selfinterested individuals [1] is important in providing insight into the mechanisms needed to support this. However, Axelrod's model suffers from significant limitations: it adopts an evolutionary approach, and assumes that information is available to all agents in the system. In particular, the model assumes that the private strategies of individuals are available to others, and that agents are omniscient in being aware of all norm violations and punishments. Because this is an unreasonable expectation, the approach does not lend itself to modelling realworld systems such as peer-to-peer networks. In response, this paper proposes alternatives to Axelrod's model, by replacing the evolutionary approach, enabling agents to learn, and by restricting the metapunishment of agents to only those where the original defection is perceived, in order to be able to apply the model to real-world domains.
In order to established a norm in a society of agents, met norms have previously been proposed as... more In order to established a norm in a society of agents, met norms have previously been proposed as a means of ensuring not that norms are complied with, but that they are enforced. Yet while experimental results have shown that metanorms are effective in fully-connected environments such as that used by Axel rod, there has been limited consideration of such metanorm models with different but more realistic topological configurations. In this paper, therefore, we consider the use of metanorms in supporting norm establishment in lattices and small world networks. Our results suggest that norm establishment is achievable in lattices and small worlds.
As forensic science and forensic statistics become increasingly sophisticated, and judges and jur... more As forensic science and forensic statistics become increasingly sophisticated, and judges and juries demand more timely delivery of more convincing scientific evidence, crime investigation is becoming progressively more challenging. In particular, this development requires more effective and efficient evidence collection strategies, which are likely to produce the most conclusive information with limited available resources. Evidence collection is a difficult task, however, because it necessitates consideration of: a wide range of plausible crime scenarios, the evidence that may be produced under these hypothetical scenarios, and the investigative techniques that can recover and interpret the plausible pieces of evidence. A knowledge based system (KBS) can help crime investigators by retrieving and reasoning with such knowledge, provided that the KBS is sufficiently versatile to infer and analyse a wide range of plausible scenarios. This paper presents such a KBS. It employs a novel compositional modelling technique that is integrated into a Bayesian model based diagnostic system. These theoretical developments are illustrated by a realistic example of serious crime investigation.
Journal of Information, Law and Technology, Jan 1, 2007
This paper describes the development of a new approach to the use of ICT for the teaching of cour... more This paper describes the development of a new approach to the use of ICT for the teaching of courses in the interpretation and evaluation of evidence. It is based on ideas developed for the teaching of science to school children, in particular the importance of models and qualitative reasoning skills. In the first part, we make an analysis of the basis of current research into "evidence scholarship" and the demands such a system would have to meet. In the second part, we introduce the details of such a system that we developed initially to assist police in the interpretation of evidence.
In the past decade, compositional modelling (CM) has established itself as the predominant knowle... more In the past decade, compositional modelling (CM) has established itself as the predominant knowledge-based approach to construct mathematical (simulation) models automatically. Although it is mainly applied to physical systems, there is a growing interest in applying CM to other domains, such as ecological and socio-economic systems. Inspired by this observation, this paper presents a method for extending the conventional CM techniques to suit systems that are fundamentally presented by interacting populations of individuals instead of physical components or processes. The work supports building model repositories for such systems, especially in addressing the most critical outstanding issues of granularity and disaggregation in ecological systems modelling.
... Concept map assessment of classroom learning: Reliability, validity, and logistical practical... more ... Concept map assessment of classroom learning: Reliability, validity, and logistical practicality. ... significance of the results to be validated, it can be flawed, especially if the validity of the ... that rely on scoring the relevance of links between concepts may be difficult to apply validly to ...
This paper is concerned with the use of compositional modelling for effectively building and (re-... more This paper is concerned with the use of compositional modelling for effectively building and (re-)using repositories of ecological models, with a focus on the important issue of model granularity. The techniques presented herein differ from existing work in that a metamodelling approach to grain choices is taken by transforming an emerging model's level of detail through disaggregation. The result is a means of choosing a model's level of detail that is sufficiently flexible for ecomodelling and that allows grain choices to be described in terms of scenario-level concepts. Model disaggregations are implemented as model transformations. Thus, different disaggregations of a model can be composed themselves. As such, a small number of disaggregations may encompass a much larger selection of models at different levels of detail. The paper shows that due to the use of dynamic constraint satisfaction techniques, grain choices are not necessarily made from the outset, but can be postponed until after an initial model is constructed.
Preference elicitation to support solving synthesis problems in certain domains (e.g. automated e... more Preference elicitation to support solving synthesis problems in certain domains (e.g. automated ecological model construction) is inhibited by a severe lack of knowledge about the criteria that motivate decision making. Yet, even in these domains, humans are able to provide some partial ordering of their preferences, based on past experience and personal opinion. Working towards an efficient representation and reasoning mechanism with such partial preference information, this paper introduces a qualitative calculus of partially ordered preferences that is rooted in order of magnitude reasoning. It then integrates this calculus in a dynamic constraint satisfaction problem. A solution algorithm for the resulting dynamic preference constraint satisfaction problem is also presented. To demonstrate the ideas, the proposed techniques are applied to sample compositional modelling and configuration tasks.
This paper presents a methodology for integrating two approaches to building decision support sys... more This paper presents a methodology for integrating two approaches to building decision support systems (DSS) for crime investigation: symbolic crime scenario abduction [16] and Bayesian forensic evidence evaluation [5]. This is achieved by means of a novel compositional modelling technique that allows for automatically generating a space of models describing plausible crime scenarios from given evidence and formally represented domain
Compositional modelling is one of the most important knowledge-based approaches to automating dom... more Compositional modelling is one of the most important knowledge-based approaches to automating domain model construction. However, its use has been limited to physical systems due to the specific presumptions made by existing techniques. Based on a critical survey of existing compositional modellers, the strengths and limitations of compositional modelling for its application in the ecological domain are identified and addressed.
A crucial aspect of evidential reasoning in crime investigation involves comparing the support th... more A crucial aspect of evidential reasoning in crime investigation involves comparing the support that evidence provides for alternative hypotheses. Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be employed for this purpose. However, the specification of BNs requires conditional probability tables describing the uncertain processes under evaluation. When these processes are poorly understood, it is necessary
Abstract Consideration of a wide range of plausible crime scenarios during any crime investigatio... more Abstract Consideration of a wide range of plausible crime scenarios during any crime investigation is important to seek convincing evidence and hence to minimise the likelihood of miscarriages of justice. It is equally important for crime investigators to be able to employ,effective and efficient evidence collection strategies that are likely to produce the most conclusive information under limited available resources.
Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be used to infer the... more Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be used to infer the probability of defence and prosecution statements based on forensic evidence. This is an important development as it helps to quantify the meaning of forensic expert testimony during court proceedings, for example, that there is "strong support" for the defence or prosecution position. Due
Crime investigation is a complex task involving vast amounts of information and requiring many di... more Crime investigation is a complex task involving vast amounts of information and requiring many different types of expert knowledge. Crime investigators would therefore benefit from the use of decision support systems to help manage this in-formation and to provide knowledge to help solve the more complex problems. Current research efforts in this area have focussed on the information management side
Probabilistic abduction extends conventional symbolic abductive reasoning with Bayesian inference... more Probabilistic abduction extends conventional symbolic abductive reasoning with Bayesian inference methods. This allows for the uncertainty underlying implications to be expressed with probabilities as well as assumptions, thus complementing the symbolic approach in situations where the use of a complete list of assumptions underlying inferences is not practical. However, probabilistic abduction has been of little use in first principle-based applications, such as abductive diagnosis, largely because no methods are available to automate the construction of probabilistic models, such as Bayesian networks (BNs). This paper addresses this issue by proposing a compositional modelling method for BNs.
Despite increasing interest in the development of intelligent techniques to aid in the prevention... more Despite increasing interest in the development of intelligent techniques to aid in the prevention and detection of crime, an important issue that has not yet been addressed by existing work is the use of knowledge based systems (KBS) to aid in the synthesis and analysis of hypothetical scenarios in major crime investigation. The main limitation of conventional KBS approaches is their lack of robustness to deal with the substantial variability of crime scenarios. This paper introduces a method to apply model based reasoning techniques to this problem. In particular, the existing compositional Bayesian modelling approach is extended and adapted to create hypothetical crime scenarios. Also, methods developed in the area of model-based diagnosis are used to support the analysis of synthesised crime scenarios.
... Cambridge University Press, 1969. 23. J. Venn. The Logic of Chance: An Essay on the Foundatio... more ... Cambridge University Press, 1969. 23. J. Venn. The Logic of Chance: An Essay on the Foundations and Province of the Theory of Probability. Macmillan, 1866. 24. J. von Plato. definetti's earliest works on the foundations of probability. Erkenntnis, 31:263-282, 1989. ...
An important cause of miscarriages of justice is the failure of crime investigators and lawyers t... more An important cause of miscarriages of justice is the failure of crime investigators and lawyers to consider important plausible explanation for the available evidence. Recent research has explored the development of decision support systems that (i) assist human crime investigators by proposing plausible crime scenarios explaining given evidence, and (ii) provide the means to analyse such scenarios. While such approaches can generate useful explanations, they are inevitably restricted by the limitations of formal abductive inference mechanisms. Building on work presented previously at this venue, this paper characterises an important class of scenarios, containing "alternative suspects" or "hidden objects", which cannot be synthesised robustly using conventional abductive inference mechanisms. The work is then extended further by proposing a novel inference mechanism that enables the generation of such scenarios.
Norms are a valuable mechanism for establishing coherent cooperative behaviour in decentralised s... more Norms are a valuable mechanism for establishing coherent cooperative behaviour in decentralised systems in which no central authority exists. In this context, Axelrod's seminal model of norm establishment in populations of selfinterested individuals [1] is important in providing insight into the mechanisms needed to support this. However, Axelrod's model suffers from significant limitations: it adopts an evolutionary approach, and assumes that information is available to all agents in the system. In particular, the model assumes that the private strategies of individuals are available to others, and that agents are omniscient in being aware of all norm violations and punishments. Because this is an unreasonable expectation, the approach does not lend itself to modelling realworld systems such as peer-to-peer networks. In response, this paper proposes alternatives to Axelrod's model, by replacing the evolutionary approach, enabling agents to learn, and by restricting the metapunishment of agents to only those where the original defection is perceived, in order to be able to apply the model to real-world domains.
In order to established a norm in a society of agents, met norms have previously been proposed as... more In order to established a norm in a society of agents, met norms have previously been proposed as a means of ensuring not that norms are complied with, but that they are enforced. Yet while experimental results have shown that metanorms are effective in fully-connected environments such as that used by Axel rod, there has been limited consideration of such metanorm models with different but more realistic topological configurations. In this paper, therefore, we consider the use of metanorms in supporting norm establishment in lattices and small world networks. Our results suggest that norm establishment is achievable in lattices and small worlds.
As forensic science and forensic statistics become increasingly sophisticated, and judges and jur... more As forensic science and forensic statistics become increasingly sophisticated, and judges and juries demand more timely delivery of more convincing scientific evidence, crime investigation is becoming progressively more challenging. In particular, this development requires more effective and efficient evidence collection strategies, which are likely to produce the most conclusive information with limited available resources. Evidence collection is a difficult task, however, because it necessitates consideration of: a wide range of plausible crime scenarios, the evidence that may be produced under these hypothetical scenarios, and the investigative techniques that can recover and interpret the plausible pieces of evidence. A knowledge based system (KBS) can help crime investigators by retrieving and reasoning with such knowledge, provided that the KBS is sufficiently versatile to infer and analyse a wide range of plausible scenarios. This paper presents such a KBS. It employs a novel compositional modelling technique that is integrated into a Bayesian model based diagnostic system. These theoretical developments are illustrated by a realistic example of serious crime investigation.
Journal of Information, Law and Technology, Jan 1, 2007
This paper describes the development of a new approach to the use of ICT for the teaching of cour... more This paper describes the development of a new approach to the use of ICT for the teaching of courses in the interpretation and evaluation of evidence. It is based on ideas developed for the teaching of science to school children, in particular the importance of models and qualitative reasoning skills. In the first part, we make an analysis of the basis of current research into "evidence scholarship" and the demands such a system would have to meet. In the second part, we introduce the details of such a system that we developed initially to assist police in the interpretation of evidence.
In the past decade, compositional modelling (CM) has established itself as the predominant knowle... more In the past decade, compositional modelling (CM) has established itself as the predominant knowledge-based approach to construct mathematical (simulation) models automatically. Although it is mainly applied to physical systems, there is a growing interest in applying CM to other domains, such as ecological and socio-economic systems. Inspired by this observation, this paper presents a method for extending the conventional CM techniques to suit systems that are fundamentally presented by interacting populations of individuals instead of physical components or processes. The work supports building model repositories for such systems, especially in addressing the most critical outstanding issues of granularity and disaggregation in ecological systems modelling.
... Concept map assessment of classroom learning: Reliability, validity, and logistical practical... more ... Concept map assessment of classroom learning: Reliability, validity, and logistical practicality. ... significance of the results to be validated, it can be flawed, especially if the validity of the ... that rely on scoring the relevance of links between concepts may be difficult to apply validly to ...
This paper is concerned with the use of compositional modelling for effectively building and (re-... more This paper is concerned with the use of compositional modelling for effectively building and (re-)using repositories of ecological models, with a focus on the important issue of model granularity. The techniques presented herein differ from existing work in that a metamodelling approach to grain choices is taken by transforming an emerging model's level of detail through disaggregation. The result is a means of choosing a model's level of detail that is sufficiently flexible for ecomodelling and that allows grain choices to be described in terms of scenario-level concepts. Model disaggregations are implemented as model transformations. Thus, different disaggregations of a model can be composed themselves. As such, a small number of disaggregations may encompass a much larger selection of models at different levels of detail. The paper shows that due to the use of dynamic constraint satisfaction techniques, grain choices are not necessarily made from the outset, but can be postponed until after an initial model is constructed.
Preference elicitation to support solving synthesis problems in certain domains (e.g. automated e... more Preference elicitation to support solving synthesis problems in certain domains (e.g. automated ecological model construction) is inhibited by a severe lack of knowledge about the criteria that motivate decision making. Yet, even in these domains, humans are able to provide some partial ordering of their preferences, based on past experience and personal opinion. Working towards an efficient representation and reasoning mechanism with such partial preference information, this paper introduces a qualitative calculus of partially ordered preferences that is rooted in order of magnitude reasoning. It then integrates this calculus in a dynamic constraint satisfaction problem. A solution algorithm for the resulting dynamic preference constraint satisfaction problem is also presented. To demonstrate the ideas, the proposed techniques are applied to sample compositional modelling and configuration tasks.
This paper presents a methodology for integrating two approaches to building decision support sys... more This paper presents a methodology for integrating two approaches to building decision support systems (DSS) for crime investigation: symbolic crime scenario abduction [16] and Bayesian forensic evidence evaluation [5]. This is achieved by means of a novel compositional modelling technique that allows for automatically generating a space of models describing plausible crime scenarios from given evidence and formally represented domain
Compositional modelling is one of the most important knowledge-based approaches to automating dom... more Compositional modelling is one of the most important knowledge-based approaches to automating domain model construction. However, its use has been limited to physical systems due to the specific presumptions made by existing techniques. Based on a critical survey of existing compositional modellers, the strengths and limitations of compositional modelling for its application in the ecological domain are identified and addressed.
A crucial aspect of evidential reasoning in crime investigation involves comparing the support th... more A crucial aspect of evidential reasoning in crime investigation involves comparing the support that evidence provides for alternative hypotheses. Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be employed for this purpose. However, the specification of BNs requires conditional probability tables describing the uncertain processes under evaluation. When these processes are poorly understood, it is necessary
Abstract Consideration of a wide range of plausible crime scenarios during any crime investigatio... more Abstract Consideration of a wide range of plausible crime scenarios during any crime investigation is important to seek convincing evidence and hence to minimise the likelihood of miscarriages of justice. It is equally important for crime investigators to be able to employ,effective and efficient evidence collection strategies that are likely to produce the most conclusive information under limited available resources.
Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be used to infer the... more Recent work in forensic statistics has shown how Bayesian Networks (BNs) can be used to infer the probability of defence and prosecution statements based on forensic evidence. This is an important development as it helps to quantify the meaning of forensic expert testimony during court proceedings, for example, that there is "strong support" for the defence or prosecution position. Due
Uploads
Papers by Jeroen Keppens