Sequencing questions to ferret out terrorists: Models and heuristics

Consider the problem of granting boarding clearance for a large population of air travelers while ferreting out any potential terrorists and denying them entry, in a short time. It is assumed that the probability of having a terrorist in the group is very small. Further assume that the process consists of asking a series of questions and the decision to clear or deny is dependent on the answer set. An efficient sequencing of the questions may reduce the number of questions needed to be asked in order to reach a decision. This problem is modeled as a question sequencing problem. The problem is intrinsically hard and hence we develop two approaches to solving the problem. The first uses a traditional greedy heuristic approach exploiting the relationship between answers and the outcome. The second adopts the decision tree approach used in classification problems to this problem. We also report on the performance of the two heuristics which does exceptionally well on problems with a very low probability of occurrence.

[1]  Vijayalakshmi Atluri,et al.  Secure interoperation for effective data mining in border control and homeland security applications , 2004, DG.O.

[2]  Henry Braun,et al.  PSYCHOMETRIC EVALUATION OF THE NEW GRE® WRITING ASSESSMENT , 2001 .

[3]  Barbara D. Klein.  Detecting errors in data: clarification of the impact of base rate expectations and incentives , 2001 .

[4]  Gang Wang,et al.  Automatically detecting deceptive criminal identities , 2004, CACM.

[5]  Hsinchun Chen,et al.  Intelligence and security informatics: information systems perspective , 2006, Decis. Support Syst..

[6]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[7]  JOHANNES GEHRKE,et al.  RainForest—A Framework for Fast Decision Tree Construction of Large Datasets , 1998, Data Mining and Knowledge Discovery.

[8]  Anand S. Kunnathur,et al.  Dynamic rescheduling of a job shop: a simulation study , 1996, Winter Simulation Conference.

[9]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[10]  Joseph S. Valacich,et al.  Emerging business models for mobile brokerage services , 2004, CACM.

[11]  Barbara D. Klein,et al.  Data quality in neural network models: effect of error rate and magnitude of error on predictive accuracy , 1999 .

[12]  H. Atabakhsh,et al.  Cross-jurisdictional criminal activity networks to support border and transportation security , 2004, Proceedings. The 7th International IEEE Conference on Intelligent Transportation Systems (IEEE Cat. No.04TH8749).

[13]  Hsinchun Chen,et al.  Using Coplink to Analyze Criminal-Justice Data , 2002, Computer.

[14]  Aiko M. Hormann,et al.  Programs for Machine Learning. Part I , 1962, Inf. Control..

[15]  Vijayalakshmi Atluri,et al.  A Credential-Based Approach for Facilitating Automatic Resource Sharing Among Ad-Hoc Dynamic Coalitions , 2005, DBSec.

[16]  Harris Wu,et al.  Mining web navigations for intelligence , 2006, Decis. Support Syst..

[17]  Jorma Rissanen,et al.  SLIQ: A Fast Scalable Classifier for Data Mining , 1996, EDBT.

[18]  Rakesh Agrawal,et al.  SPRINT: A Scalable Parallel Classifier for Data Mining , 1996, VLDB.

[19]  Donald E. Brown,et al.  An Outlier-based Data Association Method for Linking Criminal Incidents , 2003, SDM.

[20]  W. D. Linden,et al.  A maximin model for IRT-based test design with practical constraints , 1989 .

[21]  Item Banks.  A Maximin Model for Test Design with Practical Constraints , 2007 .

[22]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[23]  Ram Rachamadugu,et al.  Policies for knowledge refreshing in databases , 2009 .