IWCF07 Report Notes on the 2007 International Workshop on Computational Forensics, held in Manchester, UK, August 31st, 2007.

Katrin Franke, Cor J. Veenman and Daniel Ramos-Castro

Many improvements in forensics can be expected if recent findings in applied mathematics, statistics and computer sciences are implemented in computer-based systems. Computational methods can provide a powerful toolset for representing human expert knowledge and for implementing recognition and reasoning abilities in machines. They allow for tools that help overcome the limitation of human capabilities. Another reason for studying computational methods is that courtroom-forensic testimony is often criticized by defense lawyers as lacking a scientific basis. However, computational methods are not meant to replace the human forensic examiner.They are intended to (i) assist in basic and applied research, e.g. to establish or prove the scientific basis of a particular investigative procedure, and (ii) support the forensic examiner in his/her daily casework. Modern crime investigation shall profit from the hybrid-intelligence of humans and machines.

Computational Forensics (CF) is an emerging interdisciplinary research domain. It is understood as the hypothesis-driven investigation of a specific forensic problem using computers, with the primary goal of discovery and advancement of forensic knowledge. CF works towards (1) in-depth understanding of a forensic discipline, (2) evaluation of a particular scientific method basis and (3) systematic approach to forensic sciences by applying techniques of computer science, applied mathematics and statistics. In practice it involves modeling and computer simulation (synthesis) and/or computer-based analysis and recognition in studying and solving forensic problems.

As a scientific discipline, approaches and studies in CF need to be peer-reviewed and published for the purpose of discussion, consequent general acceptance, and rejection by the scientific community. Scientific expertise from forensics as well as computing have to be incorporated. Methods and studies have to be reviewed for their forensic and technological correctness.

The first International Workshop on Computational Forensics (IWCF) was proposed as an act of optimism and thought-out vision of a strong cross-disciplinary forum, in which novelty and discussion are paramount. Additional objectives of an IWCF were to increase awareness of the impact of computer tools in crime prevention, investigation and prosecution, and to gather widespread researchers active in various forensic, statistic or computer science disciplines.

Internationally established forensic and computational science experts supported the ambitious initiative of an IWCF. In particular their encouragement, advice and active contribution ensured the high scientific standard of the workshop held on August 31st, 2007 in Manchester, UK. Special thanks deserves the pioneering members of the international program committee: Gonzalo Álvarez Marañón, Consejo Superior de Investigaciones Científicas, Spain, Faouzi Alaya Cheikh, Gjøvik University College, Norway, Oscar Cordón, European Centre for Soft Computing, Spain, Patrick De Smet, FOD Justitie, Belgium, Zeno Geradts, Netherlands Forensic Institute, The Netherlands, Janusz Kacprzyk, Polish Academy of Sciences, Poland, Cheng-Lin Liu, Chinese Academy of Sciences, China, Milan Milosavljević, University of Belgrade, Serbia, Javier Ruiz del Solar, Universidad de Chile, Chile, Hiroshi Sako, Hitachi Central Research Laboratory, Japan, Sargur N. Srihari, University at Buffalo, USA, Lasse Øverlier, Norwegian Defence Research Establishment, Norway.

Although the 2007 IWCF was announced on a very short notice, it attracted 20 submissions from 10 countries around the world (CN, CZ, ES, IT, JP, NL, NO, SG, UK, USA). The review process has been a delicate and challenging task for the program committee, as all papers were reviewed by three experts each. In total 9 regular and 2 short papers were accepted for publication, which sum up to an acceptance rate of 55%. Covered research topics are diverse as information retrieval, data mining, digital forensics, physical trace evidence, human identification (finger print, speech recognition and handwriting), anthropology, questioned documents and forensic statistics. All accepted papers of the 2007 IWCF were published with the press of IEEE Computer Society, in particular the proceedings of the 3rd International Symposium on Information Assurance and Security (IAS 2007).

The organization of an event like IWCF is not possible without the effort and the enthusiasm of the people involved. Since the 2007 workshop was held in conjunction with the IAS 2007 special thanks deserves the general IAS chairs Ning Zhang, University of Manchester, UK, and Ajith Abraham, Norwegian University of Science and Technology, Norway who assisted with the overall organization and local arrangements.

The IWCF workshop was fashioned around dialogue, not just presentation. Each presentation was followed by an extensive discussion providing guidance and inspiration for the benefit of the researchers. In particular the different backgrounds of workshop participants supported controversy discussions from various point of view and hence scientific discourse. This stimulating atmosphere served as basis for fruitful round-table discussions on the future of computational forensics as research discipline.

The participants of the round-table discussion represent the diversity of computational forensic research, i.e., Cor J. Veenman, Netherlands Forensics Institute and University of Amsterdam, The Netherlands, Graham Leedham, University of New South Wales, Singapore, Daniel Ramos-Castro, Universidad Autonoma de Madrid, Spain, Richard Bache, University of Strathclyde, UK, Mario Refice and Donato Impedovo, Politecnico di Bari, Italy, Sergio Damas, European Centre for Soft Computing, Spain, Sargur, N. Srihari, University at Buffalo, State University of New York, USA, Katrin Franke, Norwegian Information Security Laboratory, Norway. Remote input via email was given by Colin Aitken, University of Edinburgh, UK.

Three major questions were addressed in the discussions:

  1. What should the scope of IWC 2008 be?
  2. What is the expected outcome of Computational-Forensic research?
  3. Which kind of administrative support is needed in order to establish Computational-Forensic research on a broad basis?

In the following the main outcome of the discussions is summarized that provide guidance for the further development of the field.

Add 1) CF shall not only include computer scientists. Statisticians have to be included also. In addition, there is a need of computational-science experts that do not only have expertise in analysis and evidence evaluation but also modeling and simulation, e.g. computer graphics and visualization. Further on, forensic scientists and practitioner have to be included, since their input is necessary in order to ensure that the research answers the right questions. And finally lawyers need to be included to address legal aspects of using computational methods for law enforcement. Especially, one has to take into consideration that the envisaged widespread usage of computational methods poses new legal questions, e.g. in case of false interpretations of computational results. Summarizing, the topic of CF is intrinsically multidisciplinary, and therefore contributions of forensic scientists, practitioners, layers, computer scientists, statisticians and engineers are essential.

Add 2) High-quality methods of evaluating evidence are required and good procedures for assessing the performances of the methods. Methods may not deliver final conclusions on a case but guidance in the beginning of an investigation. Computational results need to be explainable (instead of black box approaches). Most important is that computational-forensic methods may serve as decision support instead of decision taking systems. Both, methods for analysis and synthesis (modeling & simulation) are needed. These methods shall contribute establishing scientific-based and efficient work procedures that fulfill Daubert criteria. The scientific basis for a specific approach can be laid and the approach can be evaluated. Several forensic disciplines share similar demands, e.g. the proof that a trace bears sufficient individual characteristics. By means of shared knowledge, sophisticated computational methods can be efficiently adapted to a new problem domain. Error rates have to be provided. Yet, error rates are perhaps not sufficient. A proper, transparent and clear measure of goodness is the objective to report in a validity test. In the case of likelihood ratios, error rates can be misleading or incomplete. An information-theoretical method can address other performance issues, which are beyond error rates. This issue is a topic of research by itself. Further on, computational forensic research shall contribute to the evaluation and standardization of methods. Common benchmark-data sets are needed to enable the evaluation and cross-comparison of methods. Common data formats and system interfaces are also required to enable collaboration and exchange between investigation services itselves and within the scientific community. Summarizing, the outcome of CF research and development shall be (i) tools for forensic casework, (ii) scientific-based methods and efficient work procedures, and (iii) evaluation and standardization of methods and data formats.

Add 3) Several support methods are needed for CF development: international forums (e.g. conference, scientific press media) to review and exchange research results, education and training to prepare current and future researchers and practitioners, and financial support for research and development.

In summary: Computational Forensics holds the potential to greatly benefit all of the forensic sciences. For the computer scientist it poses a new frontier where new problems and challenges are to be faced. The potential benefits to society, meaningful inter-disciplinary research, and challenging problems should attract high quality students and researchers to the field.

October 16th, 2007