The application of AI to law

There is much interest in moving AI out into real world applications, a move which has been encouraged by recent funding which has attempted to show industry and commerce can benefit from the Fifth Generation of computing. In this article I suggest that the legal application area is one which is very much more complex than it might — at first sight — seem. I use arguments from the sociology of law to indicate that the viewing of the legal system as simply a rule-bound discipline is inherently nave. This, while not new in jurisprudence, is — as the literature of AI and law indicates — certainly novel to the field of artificial intelligence. The socio-legal argument provided is set within the context of AI as one more example of the failure of scientific success and method to easily transmit itself over into the social sciences.     

Optimum design of nonlinear structures with path dependent reponse

Optimum design of structures with path dependent response is studied in this paper. The direct differentiation and the adjoint structure methods of design sensitivity analysis are summarized. The reference volume concept is used to unify the conventional and shape design problems. It is concluded that the direct differentiation method is more effective for this class of problems. The design sensitivity analysis — developed with continuum formulation — is discretized using the finite element method. Two cases for an example problem are optimized using a sequential quadratic programming algorithm to demonstrate how the developed procedures work and to study the optimization process for the problems with path dependent response.     

Non-conventional matrix calculus in the analysis of rank deficient Hankel matrices of finite dimensions

New algebraic results are obtained that describe the relation between the structure of rank deficient Hankel matrices of finite dimensions and their singular value decomposition. Use is made of the ‘non-conventional’ matrix calculus of Khatri—Rao and Kronecker products. One of the important results is the parametrization of all rank deficient Hankel matrices of finite dimensions that share (a subset of) the same set of minimal system poles. The results are illustrated with a clarifying example.     

Unidirectional loop network layout by a LP heuristic and design of telecommunications networks

In this paper, we present a new linear programming-based heuristic procedure for optimal design of the unidirectional loop network layout problem. The heuristic procedure employs a linear programming formulation and solves the problem using the flow matrix of the unidirectional loop problem. To find an optimal solution, one can either generate all possible solutions or use a branch-and-bound procedure. But, both above methods require very high computational time and computer memory for larger problems. The heuristic developed in this paper is quite fast and obtains near optimal solutions. The heuristic procedure was tested on 16 different problems selected from the literature. The results showed that in most cases optimal—and in a few cases near optimal—solutions were obtained with very little computational time. Several examples are discussed. We also demonstrate that the above problem formulation and approach can be used to solve a special class of telecommunication networks where a set of computers (or processors) are attached by unidirectional point-to-point links around a loop.     

A convex characterization of distributed control problems in spatially invariant systems with communication constraints

In this paper we consider the problem of distributed controller design in spatially invariant systems for which communication among sites is limited. In particular, the controller is constrained so that information is propagated with a delay that depends on the distance between subsystems—a structure we refer to as “funnel” causality. We show that the problem of optimal design can be cast as a convex problem provided that the plant has a similar funnel-causality structure, and the propagation speeds in the controller are at least as fast as those in the plant. As an example, we consider the case of the wave dynamics with limited propagation speed control.     

An Irreversible Economic Approach to the Theory of Production

The general equilibrium theory — the most elaborated form of the value theory — is unable to integrate firms unless under the etiquette of 'techniques'. However, to study important economic questions, as for example growth, without firms is at most doubtful. As an alternative to the attempts aiming to rehabilitate firms (as for example Coase and the theory of organization) we propose to adopt a new approach which not only treats firms as real decision makers, i.e. they make irreversible economic decisions, but describes them from a dynamic point of view, i.e. an Onsagerian type of dynamics is developed. The main charateristics of this approach are the no-loss choice rule vis-à-vis the profit maximization, and the integration of quantity signals (and past experience) vis-à-vis the unique price signals. As a result, we are capable of separating three questions which are generally not distinguished in neoclassical economics, namely the evaluation (profitability), the choice of production level and that of trade. Hence, even in the linear technology and price taker case we have a meaningful answer as, for instance, the level of output.     

What is a “Good” Encoding of Guarded Choice?

The π-calculus with synchronous output and mixed-guarded choices is strictly more expressive than the π-calculus with asynchronous output and no choice. This result was recently proved by C. Palamidessi and, as a corollary, she showed that there is no fully compositional encoding from the former into the latter that preserves divergence-freedom and symmetries. This paper argues that there are nevertheless “good” encodings between these calculi. In detail, we present a series of encodings for languages with (1) input-guarded choice, (2) both input- and output-guarded choice, and (3) mixed-guarded choice, and investigate them with respect to compositionality and divergence-freedom. The first and second encoding satisfy all of the above criteria, but various “good” candidates for the third encoding—inspired by an existing distributed implementation—invalidate one or the other criterion. While essentially confirming Palamidessi's result, our study suggests that the combination of strong compositionality and divergence-freedom is too strong for more practical purposes.     

Policy challenges in building dependability in global infrastructures

[This paper was given at Compsec 2002, in London, on 30 October 2002].Global or continental critical infrastructures — including electric power, telecommunications, and the Internet — are now the control plane for advanced economies. The occasional failures of these key infrastructures illustrate not only our dependence, but also the unanticipated interdependencies between systems. For example, the 1998 failure of a single telecommunications satellite, Galaxy 4, led to an outage of nearly 90% of all pagers in the United States, while also causing a number of unanticipated failures: many banking and financial services (credit card purchases, automated teller machines) were interrupted, as was communications with doctors and emergency workers [1].With awareness of economic and social dependence on these distributed infrastructures has come a growing concern about their reliability and security. Defense against deliberate attack — critical infrastructure protection — emerged as part of the US national security posture in the mid-1990s with the work of the President’s Commission on Critical Infrastructure, and was codified by Presidential Decision Directive 63 in 1998. Other nations are also beginning to develop national strategies for infrastructure protection.Reliability is more than protection against deliberate attack. An accidental cut of a fiber optic trunk shut down air traffic control along the east coast of the US for a day. A cascading series of events, starting with a tree limb falling, caused much of the western US to lose electricity.The challenge of improving the reliability of global networked infrastructures presents us with significant analytical and decision-making complexities, with both technical and policy relevant dimensions [2]. This paper — using principally examples from the Internet and other distributed IT systems — presents two perspectives on these complexities. First is to present critical global infrastructures as complex adaptive systems, which share certain characteristics that policy makers and managers need to account for. Secondly, the balance of the paper outlines five major dimensions of the analytical and decision-making complexity, and presents the research and policy-making agendas that need to be addressed if we are to significantly improve the reliability of global infrastructures.Neither of these perspectives is purely technical or engineering based. Success in increasing the reliability of global infrastructures will require much more analytically sophisticated research in, among other topics, the issue areas identified here — in addition to ongoing technology-based research.     

Navigational techniques to improve the display of large two-dimensional spaces

Often the components of a problem can be arrayed on a two-dimensional information space—for example, as an abstract tree or hypertext—far too large to fit onto a computer display. With current navigational techniques it is often difficult for users to keep track of their location or to move rapidly to remote locations in the space. We implemented two similar direct-manipulation techniques, both of which use a map window—a miniature of the entire information space—with a wire-frame box to aid users in remembering their location. The first technique allows the user to rapidly roam over the information space by moving the location of the wire-frame box. The second allows for zooming as well as roaming. A controlled experiment compared the above techniques to scroll bars for determining whether a target word was in a large balanced binary tree of words. The experiment also examined the merit of the map window. Map windows significantly improved user performance, and the roam and zoom techniques were found to be significantly faster than the scroll bars. Our observation of subjects and their verbal protocol indicated that a paging feature on the scroll bars—allowing rapid movement of the screen a page in any direction—was advantageous.     

Design for a fax

We argue that button-controlled devices can be made better to improve their usability, and that there are routine ways to do so that can be effectively employed in the early stages of the design process.By way of example, we examine the design of a 1993/4 domestic phone/fax/answerphone machine, and show that alternative designs have quantitative advantages over the original. We point to many strange — obscure, undocumented, non-functional — features, and raise questions about the process that led to the design, and what role either human factors or computer science played in it.     

A scientific psychologistic foundation for theories of meaning

I propose, develop and defend the view that theories of meaning — for instance, a theory specifying the logical form or truth conditions of natural language sentences — should be naturalized to scientific psychological inquiry. This involves both psychologism — the claim that semantics characterizes psychological states — and scientific naturalism — the claim that semantics will depend on the data and theories of scientific psychology. I argue that scientific psychologism is more plausible than the traditional alternative, the view that a theory of meaning is a priori. After defending scientific psychologism against several objections, I offer a speculative proposal that shows how a theory of meaning can be integrated into scientific psychology.This essay is a descendant of part of my MIT doctoral dissertation. Thanks to Ned Block, Noam Chomsky, Jim Higginbotham and Bob Stalnaker for their comments and criticisms on earlier versions, and for helpful discussion of the issues.     

On Randomization in On-Line Computation

This paper concerns two fundamental but somewhat neglected issues, both related to the design and analysis of randomized on-line algorithms. Motivated by early results in game theory we define several types of randomized on-line algorithms, discuss known conditions for their equivalence, and give a natural example distinguishing between two kinds of randomizations. In particular, we show thatmixedrandomized memoryless paging algorithms can achieve strictly better competitive performance thanbehavioralrandomized algorithms. Next we summarize known—and derive new—“Yao principle” theorems for lower bounding competitive ratios of randomized on-line algorithms. This leads to four different theorems for bounded/unbounded and minimization/maximization problems.     

The analysis of solids without mesh generation using trimmed patch boundary elements

The tools available to the mechanical engineer—for example, geometric modeling and computer-aided analysis—are individually quite powerful, but they are based on different geometric representations. Hence, they do not always work well together. In this paper an analysis method is presented that operates directly on the geometric modeling representation. Therefore, the time-consuming and error-prone procedure of generating a mesh is skipped. The method is based on boundary integral equations, but unlike previously published methods, the boundary elements aren-sided trimmed patches, the same patches that are used by modern geometric modelers to represent complex solids. The method is made practical by defining shape functions over the trimmed patches in such a way that the number of degrees of freedom can be controlled. This is done by using a concept called virtual nodes. The paper begins by deriving the trimmed patch boundary element. Then its properties are discussed in comparison with existing boundary element and finite element methods, and several examples are given.     

System Embedding. Control with Reduced Observer

Two interrelated problems—design of the reduced observer of plant state separately and together with its control system—were considered from the standpoint of designing the multivariable linear systems from the desired matrix transfer functions. The matrix equations defining the entire constructive class of solutions of the posed problems were obtained using the system embedding technology. As was demonstrated, control based on the reduced observer is capable to provide the desired response to the control input, as well as the response to the nonzero initial conditions, only for the directly measurable part of the components of the state vector. An illustrative example was presented.     

Effective Strategies of Forming Scenarios of the Behavior of Complex Systems in the ACSES's

Theoretical results are set out for the development of bases of the scenario calculus evolved in the framework of design of an automatic control system of extreme situations (ACSES's). Stages are considered of the process of optimization of stepwise strategies for the formulation of scenarios: the accounting of the uncertainty and the functional optimization, which re ect specific features of the analysis of scenarios and the process of making decisions in the ACSES's. The notions of the above stages are formalized and examples are given. Their generalized characteristics—the measure of optimism and the measure of practicability of working out a scenario—are introduced.     

Semiactive control of a vibrating system by means of electrorheological fluids

A control scheme is designed for the purpose of suppression of vibratory motion of a dynamical system. The efficacy and robustness of the controller vis a vis unknown but bounded disturbances and state measurement errors is investigated analytically and numerically. As an example of a dynamical system we consider a single degree of freedom mass—spring—damper system that is excited by an unknown force. The control scheme presupposes that the spring and damping coefficients can be varied within prescribed bounds, albeit not independently. The construction of such a semiactive controller can be realized by using the properties of so calledelectrorheological fluids (see [1] for relevant experimental investigations). The called for changes in spring and damping properties can be effected in microseconds since the control does not involve the separate dynamics (inertia) of usual actuators. The design of the controller is based on Lyapunov stability theory which is also utilized to investigate the stabilizing properties of the controller. To accommodate state measurement errors the proposed control scheme is combined with afuzzy control concept. Simulations are carried out for examples of periodic, continuous nonperiodic, discontinuous periodic and random excitation forces.     

On-Line Routing of Real-Time Messages

The problem of routing unit-length, real-time messages in a distributed system is considered. An on-line routing algorithm is one that routes messages without any knowledge of future arrivals of messages. An on-line algorithm is said to be optimal if it produces a feasible route whenever one exists. In this article, we study the issue whether it is possible to have an optimal on-line algorithm for the following networks—unidirectional ring, out-tree, in-tree, bidirectional tree, and bidirectional ring. The problem is considered under various restrictions of the four parameters—origin node, destination node, release time, and deadline. We show that: (1) for a unidirectional ring, no such algorithm can exist unless one of the four parameters is fixed (i.e., all messages have identical values for that parameter); (2) for an out-tree, no such algorithm can exist unless one of the three parameters—origin node, destination node, and release time—is fixed; (3) For an in-tree, no such algorithm can exist unless one of the three parameters—origin node, destination node, and deadline —is fixed; (4) for a bidirectional tree, no such algorithm can exist unless the origin node or the destination node is fixed; (5) for a bidirectional ring, no such algorithm can exist unless the origin node and either the destination node or the release time are fixed. Our results give a sharp boundary delineating those instances for which an optimal algorithm exists and those for which no such algorithm can exist.     

Automatic analysis of geologic structure from dip-strike data

A new technique for structural analysis is described. From dip-strike data, a dip—strike distribution map, structure contour map, and block diagram are automatically prepared on an X-Y plotter by using four computer programs. Three test examples of theoretical data are treated as well as an actual example. The Izumi Group (Upper Cretaceous) is revealed to have a more complicated structure than previously recognized.     

Conditional quantum channel and its certain construction

This paper presents a sophisticated concept and a construction method of conditional linear isometric operators which represent optical information channels belonging to a kind of non-unitary process — the conditional quantum channel. The conditional linear isometric representations of quantum channel are studied as an example in quantum optics. It is shown that the concept of conditional linear isometry is very natural.     

Simulation aided design of organizational structures in manufacturing systems using structuring strategies

This paper presents a simulation aided approach for designing organizational structures in manufacturing systems. The approach is based on a detailed modeling and characterization of the forecasted order program, especially of elementary processes, activity networks and manufacturing orders. Under the use of the organization modeling system FORM, that has been developed at the ifab-Institute of Human and Industrial Engineering of the University of Karlsruhe, structuring strategies—e.g., a process-oriented strategy—can be applied in order to design organizational structures in manufacturing systems in a flexible and efficient way. Following that, a dynamical analysis of the created manufacturing structures can be carried out with the simulation tool FEMOS, that has also been developed at the ifab-Institute. The evaluation module of FEMOS enables to measure the designed solutions with the help of logistical—e.g., lead time degree—and organizational—e.g., degree of autonomy—key data. This evaluation is the basis for the identification of effective manufacturing systems and also of improvement potentialities. Finally, a case study is presented in this paper designing and analyzing different organizational structures of a manufacturing system where gear boxes and robot grip arms were manufactured.