Classifying nodes in networks is a task with a wide range of applications. It can be particularly useful in anomaly and fraud detection. Many resources are invested in the task of fraud detection due to the high cost of fraud, and being able to automatically detect potential fraud quickly and precisely allows human investigators to work more efficiently. Many data analytic schemes have been put into use; however, schemes that bolster link analysis prove promising. This work builds upon the belief propagation algorithm for use in detecting collusion and other fraud schemes. We propose an algorithm called SNARE (Social Network Analysis for Risk Evaluation). By allowing one to use domain knowledge as well as link knowledge, the method was very successful for pinpointing misstated accounts in our sample of general ledger data, with a significant improvement over the default heuristic in true positive rates, and a lift factor of up to 6.5 (more than twice that of the default heuristic). We also apply SNARE to the task of graph labeling in general on publicly-available datasets. We show that with only some information about the nodes themselves in a network, we get surprisingly high accuracy of labels. Not only is SNARE applicable in a wide variety of domains, but it is also robust to the choice of parameters and highly scalable-linearly with the number of edges in a graph.

In recent years, there have been several large accounting frauds where a company's financial results have been intentionally misrepresented by billions of dollars. In response, regulatory bodies have mandated that auditors perform analytics on detailed financial data with the intent of discovering such misstatements. For a large auditing firm, this may mean analyzing millions of records from thousands of clients. This paper proposes techniques for automatic analysis of company general ledgers on such a large scale, identifying irregularities - which may indicate fraud or just honest errors - for additional review by auditors. These techniques have been implemented in a prototype system, called Sherlock, which combines aspects of both outlier detection and classification. In developing Sherlock, we faced three major challenges: developing an efficient process for obtaining data from many heterogeneous sources, training classifiers with only positive and unlabeled examples, and presenting information to auditors in an easily interpretable manner. In this paper, we describe how we addressed these challenges over the past two years and report on experiments evaluating Sherlock.

A system called Merle-Soar is described which demonstrates how a specific architecture for general intelligence and learning (Soar) can be used to reduce scheduling effort when solving simple scheduling problems. In particular, we describe how Merle-Soar schedules sequences of jobs on a single bottleneck machine in a job shop. The knowledge of dispatching, acquired from examining how a human expert performs the task, is cast as search rules. A study was conducted which examined the extent to which learning could contribute to decreases in scheduling effort; specifically, the contribution of learning within-tasks was explored—the change in reasoning effort while solving a particular scheduling problem as knowledge is accumulated from successive trials. The results indicated that dramatic reductions in scheduling effort (in terms of the Soar architecture) were obtained. Knowledge gained early in the scheduling task was subsequently applied later in the task to reduce deliberation, and knowledge gained from one trial successfully reduced deliberation effort in subsequent trials. Additionally, the reduction exhibited the general power law of learning documented in psychological studies of skill acquisition. INFORMS Journal on Computing, ISSN 1091-9856, was published as ORSA Journal on Computing from 1989 to 1995 under ISSN 0899-1499.

The starting point for this commentary is Sacks’ and Doyle's conclusion that a central problem for qualitative physics is automating mathematical model formulation. We believe that model formulation is also a central problem for operations research, and although we have focused on models for production planning rather than for engineering systems analysis, our experience confirms that of Sacks and Doyle, that at least parts of model formulation are amenable to automation. In terms of their recommendations for future research, their strategy seems to emphasize the formalization of mathematical knowledge. We wish to stress that understanding the design or analysis context, the problem domain, and resource constraints on the modeling process is equally important. Methods used in cognitive psychology for understanding human problem solving, such as protocol analysis, can complement mathematical study by helping us understanding the processing that human modelers use to bring mathematical knowledge to bear. We have been using the results of such analyses to guide the creation of a model formulation system (MFS) within the Soar architecture. The use of cognitive studies and computer models in tandem seems to represent a viable strategy for making progress in this area.

The analysis and execution of partial algorithm descriptions is an important part of the algorithm design process (as is borne out by studying the behavior of human algorithm designers). In this paper, we describe a language for representing partially designed algorithms and a process, developmental evaluation, that can discover useful knowledge to guide design. Using these and other results from our research in artificial intelligence, we are building a system, DESIGNER, that automatically designs algorithms. This paper also compares developmental evaluation to execution and analysis techniques used for testing complete programs and for validation of abstract specifications; concepts similar to those found in developmental evaluation are thus shown to apply to all stages of the software life cycle.