Constraint Satisfaction Methods Research Articles

Visual analogy: Viewing analogical retrieval and mapping as constraint satisfaction problems

The core issue of analogical reasoning is the transfer of relational knowledge from a source case to a target problem. Visual analogical reasoning pertains to problems containing only visual knowledge. Holyoak and Thagard proposed that the retrieval and mapping tasks of analogy in general can be productively viewed as constraint satisfaction problems, and provided connectionist implementations of their proposal. In this paper, we reexamine the retrieval and mapping tasks of analogy in the context of diagrammatic cases, representing the spatial structure of source and target diagrams as semantic networks in which the nodes represent spatial elements and the links represent spatial relations. We use a method of constraint satisfaction with backtracking for the retrieval and mapping tasks, with subgraph isomorphism over a particular domain language as the similarity measure. Results in the domain of 2D line drawings suggest that at least for this domain the above method is quite promising.

Generating customized yet factually consistent information: a constraint satisfaction approach

We present an information customization framework that leverages a hybrid of adaptive hypermedia and intelligent techniques, in particular constraint satisfaction methods, to generate customized and factually consistent information based on a user profile. Information customization is modeled as a constraint satisfaction problem, whereby a solution is derived by (a) satisfying user-model constraints to select a user-specific set of ‘information snippets’; and (b) establishing inter-snippet consistency to ensure that all snippets are compatible with each other. Our approach takes the unique step of establishing factually consistency – via the satisfaction of inter-snippet constraints – between heterogeneous information snippets. A customized information package is generated by systematically synthesizing the set of user-specific and factually consistent information snippets. The featured information customization framework incorporates variations of various search and constraint satisfaction methods. The work is applied in an E-Healthcare setting leading to the generation of customized healthcare information.

Design of Density Functionals by Combining the Method of Constraint Satisfaction with Parametrization for Thermochemistry, Thermochemical Kinetics, and Noncovalent Interactions.

We present a new hybrid meta exchange-correlation functional, called M05-2X, for thermochemistry, thermochemical kinetics, and noncovalent interactions. We also provide a full discussion of the new M05 functional, previously presented in a short communication. The M05 functional was parametrized including both metals and nonmetals, whereas M05-2X is a high-nonlocality functional with double the amount of nonlocal exchange (2X) that is parametrized only for nonmetals. In particular, M05 was parametrized against 35 data values, and M05-2X is parametrized against 34 data values. Both functionals, along with 28 other functionals, have been comparatively assessed against 234 data values: the MGAE109/3 main-group atomization energy database, the IP13/3 ionization potential database, the EA13/3 electron affinity database, the HTBH38/4 database of barrier height for hydrogen-transfer reactions, five noncovalent databases, two databases involving metal-metal and metal-ligand bond energies, a dipole moment database, a database of four alkyl bond dissociation energies of alkanes and ethers, and three total energies of one-electron systems. We also tested the new functionals and 12 others for eight hydrogen-bonding and stacking interaction energies in nucleobase pairs, and we tested M05 and M05-2X and 19 other functionals for the geometry, dipole moment, and binding energy of HCN-BF3, which has recently been shown to be a very difficult case for density functional theory. We tested eight functionals for four more alkyl bond dissociation energies, and we tested 12 functionals for several additional bond energies with varying amounts of multireference character. On the basis of all the results for 256 data values in 18 databases in the present study, we recommend M05-2X, M05, PW6B95, PWB6K, and MPWB1K for general-purpose applications in thermochemistry, kinetics, and noncovalent interactions involving nonmetals and we recommend M05 for studies involving both metallic and nonmetallic elements. The M05 functional, essentially uniquely among the functionals with broad applicability to chemistry, also performs well not only for main-group thermochemistry and radical reaction barrier heights but also for transition-metal-transition-metal interactions. The M05-2X functional has the best performance for thermochemical kinetics, noncovalent interactions (especially weak interaction, hydrogen bonding, π···π stacking, and interactions energies of nucleobases), and alkyl bond dissociation energies and the best composite results for energetics, excluding metals.

Optimization of 3G mobile network design using a hybrid search strategy

This paper proposes an efficient constraint-based optimization model for the design of 3G mobile networks, such as universal mobile telecommunications system (UMTS). The model concerns about finding a set of sites for locating radio network controllers (RNCs) from a set of pre-defined candidate sites, and at the same time optimally assigning node Bs to the selected RNCs. All these choices must satisfy a set of constraints and optimize an objective function. This problem is NP-hard and consequently cannot be practically solved by exact methods for real size networks. Thus, this paper proposes a hybrid search strategy for tackling this complex and combinatorial optimization problem. The proposed hybrid search strategy is composed of three phases: A constraint satisfaction method with an embedded problem-specific goal which guides the search for a good initial solution, an optimization phase using local search algorithms, such as tabu algorithm, and a post-optimization phase to improve solutions from the second phase by using a constraint optimization procedure. Computational results show that the proposed search strategy and the model are highly efficient. Optimal solutions are always obtained for small or medium sized problems. For large sized problems, the final results are on average within 5.77% to 7.48% of the lower bounds.

Modeling Uncertainty in Clinical Diagnosis Using Fuzzy Logic

This paper describes a fuzzy approach to computer-aided medical diagnosis in a clinical context. It introduces a formal view of diagnosis in clinical settings and shows the relevance and possible uses of fuzzy cognitive maps. A constraint satisfaction method is introduced that uses the temporal uncertainty in symptom durations that may occur with particular diseases. The method results in an estimate of the stage of the disease if the temporal constraints of the disease in relation to the occurrence of the symptoms are satisfied. A lightweight fuzzy process is described and evaluated in the context of diagnosis of two confusable diseases. The process is based on the idea of an incremental simple additive model for fuzzy sets supporting and negating particular diseases. These are combined to produce an index of support for a particular disease. The process is developed to allow fuzzy symptom information on the intensity and duration of symptoms. Results are presented showing the effectiveness of the method for supporting differential diagnosis.

Constraint Satisfaction Methods for Solving the Staff Transfer Problem

The Staff Transfer Problem is concerned with the assignment of transfer postings to employees in large organizations. Staff transfers are an important issue in Human Resource Management in countries like India and China that have many large public sector undertakings. The Staff Transfer Problem can be viewed as a Constraint Satisfaction Problem, and methods such as Simulated Annealing, Genetic Algorithms, Satisfiability (GSAT), and Conflict Directed Backjumping can all be employed to solve randomly generated problem instances. Computer experiments indicate that Simulated Annealing is the best method of solution. GSAT with a tabu list yields solutions of good quality but is unable to solve large instances. Genetic Algorithms is also good, but it takes much more time than Simulated Annealing, Conflict Directed Backjumping, a deterministic search technique, is markedly inferior to the other methods. Thus for solving the Staff Transfer Problem, randomized approaches appear to be superior to deterministic ones.

The cyclic genetic code as a constraint satisfaction problem

We consider in this paper a combinatorial problem we propose to solve with a constraint satisfaction method well defined at the beginning of the real game of life: this problem consists in extracting from the classical degenerate genetic code made of 64 triplets, a non-degenerate cyclic code made of 22 triplets, which we call the cyclic AB code. It is made up of the following chain: AUGGUGCCAUUCAAGACUAUGA, where the letters A, U, C, G represent the classical symbols of the (purine and pyrimidine) bases of the genetic code. The chain in a circular form presents the following features: (1) It begins with the initiation codon AUG and ends with the termination codon UGA. When it is in cyclic form, it can be read triplet after triplet by choosing one and only one representative of each synonymy class in the classical degenerate genetic code. The chain, therefore, possesses one and only one codon for each amino-acid. (2) Except for the pair CG, the chain contains the 15 other possible pairs of bases which start the triplets (satisfying the classical ”wobble” hypothesis). (3) It corresponds to the major part of the loops of the present Oenothera Gly-tRNA. (4) It contains the most frequent triplets, but not the most rare ones, appearing in the genome of numerous species. This structure, also found in the loops part of the tRNAs, contains an excess of A, U bases with respect to G, C bases, like in the present tRNAs. More, its 3D structure has strong similarities with a precise part of the 3D shape of the present tRNAs. Therefore, we propose the cyclic AB code, solution of a realistic constraint satisfaction problem, as a primitive genetic structure which presents the essential coding properties of the present genetic code and has the primitive function of favouring (like the present tRNAs) the formation of peptides and proteins.

Computer aided fuzzy medical diagnosis

This paper describes a fuzzy approach to computer aided medical diagnosis in a clinical context. It introduces a formal view of diagnosis in clinical settings and shows the relevance and possible uses of fuzzy cognitive maps and fuzzy logic. A constraint satisfaction method is introduced which uses the temporal uncertainty in symptom durations that may occur with particular diseases. Together with fuzzy symptom descriptions, the method results in an estimate of the stage of a disease if the temporal constraints of the disease in relation to the occurrence of the symptoms are satisfied. The approach is evaluated through simulation experiments showing the effects of symptom ordering, temporal uncertainty and symptom strengths on the diagnosis efficiency. The method is effective and can be developed further using second order (Type 2) fuzzy logic to better represent uncertainty in the clinical context thus improving differential diagnosis accuracy.

Potential carbon mitigation and income in developing countries from changes in use and management of agricultural and forest lands.

The many opportunities for mitigating atmospheric carbon emissions in developing countries include reforesting degraded lands, implementing sustainable agricultural practices on existing lands and slowing tropical deforestation. This analysis shows that over the next 10 years, 48 major tropical and subtropical developing countries have the potential to reduce the atmospheric carbon burden by about 2.3 billion tonnes of carbon. Given a central price of $10 per tonne of carbon and a discount rate of 3%, this mitigation would generate a net present value of about $16.8 billion collectively for these countries. Achieving these potentials would require a significant global effort, covering more than 50 million hectares of land, to implement carbon-friendly practices in agriculture, forest and previously forested lands. These estimates of host-country income potentials do not consider that outside financial investment may or may not be available. Our calculations take no account of the additional benefits of carbon sequestration in forest soils undergoing reforestation, increased use of biomass and reduced use of fossil-fuel inputs and reduced agricultural emissions. In all events, realizing these incomes would necessitate substantially greater policy support and investment in sustainable land uses than is currently the case.

Extracting knowledge-based energy functions from protein structures by error rate minimization: Comparison of methods using lattice model

We describe a general framework for extracting knowledge-based energy function from a set of native protein structures. In this scheme, the energy function is optimal when there is least chance that a random structure has a lower energy than the corresponding native structure. We first show that subject to certain approximations, most current database-derived energy functions fall within this framework, including mean-field potentials, Z-score optimization, and constraint satisfaction methods. We then propose a simple method for energy function parametrization derived from our analysis. We go on to compare our method to other methods using a simple lattice model in the context of three different energy function scenarios. We show that our method, which is based on the most stringent criteria, performs best in all cases. The power and limitations of each method for deriving knowledge-based energy function is examined.

Constraint Satisfaction Methods for Applications in Engineering

Constraints provide declarative descriptions of important requirements related to engineering projects. Most existing algorithms for constraint satisfaction require input consisting of binary constraints on variables that have discrete values. Such restrictions limit their use in engineering since complex constraints, involving several variables that have discrete and numeric values, are common. This paper provides an approach for decision support through approximating solution spaces that are defined by constraints. Our algorithm is not limited to a specific type of constraint, but handles numeric and discrete variables in the same frame-work. Since a new type of local consistency narrows down the search space effectively, full-scale engineering tasks, such as designs involving hundreds of variables, are accommodated without excessive computational complexity. The approach is demonstrated for selection of appropriate wind bracing for single story steel-framed buildings. Results may be used for input into other tools containing algorithms such as those offering (i) higher levels of consistency, (ii) optimally directed point-solution search, and (iii) simulation behaviour. Finally, extension to dynamic constraint satisfaction using different combinations of activation conditions is straightforward. It is expected that this approach will improve the performance of many existing and future computeraided engineering tools.

A scheme for unifying optimization and constraint satisfaction methods

Optimization and constraint satisfaction methods are complementary to a large extent, and there has been much recent interest in combining them. Yet no generally accepted principle or scheme for their merger has evolved. We propose a scheme based on two fundamental dualities: the duality of search and inference, and the duality of strengthening and relaxation. Optimization as well as constraint satisfaction methods can be seen as exploiting these dualities in their respective ways. Our proposal is that rather than employ either type of method exclusively, one can focus on how these dualities can be exploited in a given problem class. The resulting algorithm is likely to contain elements from both optimization and constraint satisfaction, and perhaps new methods that belong to neither.

Exploring use of multiobjective genetic algorithms for constrained process scheduling

Abstract Chemical processing offers constrained, complex scheduling problems (Nott and Lee, 1997). Genetic algorithms (GAs) are evolutionary-based optimisation tools (Holland, 1975), suited to schedule optimisation search, (Davis, 1985). Multi-objective GAs (MOGA) allow multiple objective optimisation (Fonseca and Fleming, 1998), as well as a technique for constraint handling, and the exploration of the inherent trade-offs between goals and constraints. A batch process problem is introduced, and a MOGA is applied to a plant model. The MOGA acts as a flexible optimisation tool, incorporating the framework of the ISA standard S88 (Love and Bunch, 1998) and allowing the user to specify plant variables. This paper explores the MOGA as an optimisation and constraint satisfaction method.

Advanced, constrained, genetic algorithm load flow method

An efficient and reliable constrained genetic algorithm based load flow algorithm is developed. The components and the constraint satisfaction methods in the algorithm are presented. A nodal voltage differential technique and a gradient method are developed to accelerate the solution process in the algorithm. A method for dealing with generator reactive power limits is also developed. The new algorithm can determine multiple load flow solutions and can be used to determine the loadability limits of transmission systems. Its power is demonstrated through its applications to evaluate the loadability of the IEEE 30-, 57- and 118-node systems. Detailed results on the 30-node system are presented.

Distributed Constraint Satisfaction in NEDIP

This article presents a negotiation scheme as extension to the distributed planning system NEDIP (NEgotiation based DIstributed Planning) that runs on a set of workstations. In this negotiation scheme constraints are used in order to model a typical negotiation problem of several autonomous agents. The algorithm presented is a hybrid approach using message exchange and constraint satisfaction methods.

Diagrammatic Transformation Processes on Two-dimensional Relational Maps

Research on qualitative spatial reasoning has produced an impressive number of systems of relations for representing topological, ordinal and metrical information. In comparison, the number of reasoning procedures that has been proposed is rather limited. Most approaches simply rely on standard constraint propagation techniques such as the path consistency algorithm. Evidence from cognitive science indicates that human reasoners solve spatial relational inferences not by constraint satisfaction methods but by constructing and inspecting mental models in visuo-spatial working memory. From a computational point of view, this strategy combines the relational representations used in qualitative spatial reasoning with the local spatial transformations studied in diagrammatic reasoning. However, applying local transformations to arbitrary relational representations involves solving the intractable subgraph isomorphism problem. The paper describes a class of representations, relational maps, for which the problem becomes tractable and for which, as a consequence, diagrammatic inference is implementable by efficient local transformations.

Automated analysis of protein NMR assignments using methods from artificial intelligence

An expert system for determining resonance assignments from NMR spectra of proteins is described. Given the amino acid sequence, a two-dimensional 15N- 1H heteronuclear correlation spectrum and seven to eight three-dimensional triple-resonance NMR spectra for seven proteins, AUTOASSIGN obtained an average of 98% of sequence-specific spin-system assignments with an error rate of less than 0.5%. Execution times on a Sparc 10 workstation varied from 16 seconds for smaller proteins with simple spectra to one to nine minutes for medium size proteins exhibiting numerous extra spin systems attributed to conformational isomerization. AUTOASSIGN combines symbolic constraint satisfaction methods with a domain-specific knowledge base to exploit the logical structure of the sequential assignment problem, the specific features of the various NMR experiments, and the expected chemical shift frequencies of different amino acids. The current implementation specializes in the analysis of data derived from the most sensitive of the currently available triple-resonance experiments. Potential extensions of the system for analysis of additional types of protein NMR data are also discussed.

On computing logic programs

In this paper we present and compare some classical problem-solving methods for computing the stable models of logic programs with negation. Using a graph theoretic representation of logic programs and their stable models, we discuss and compare linear programming, propositional satisfiability, constraint satisfaction, and graph methods.

Automated analysis of nuclear magnetic resonance assignments for proteins

Recent developments in protein NMR technology provide spectral data that are highly amenable to analysis by computer software systems. Automated methods of analysis use constraint satisfaction, pseudoenergy minimization, directed search, neural net, simulated annealing, and/or genetic algorithms to establish sequential links and sequence-specific assignments. The most advanced systems provide automated analysis of complete backbone and extensive side-chain resonance assignments for proteins of 50–150 amino acids.

Automated Location of Peak Patterns in Multidimensional NMR Spectra: A Tool for Resonance Assignment in Proteins

Recent developments in protein NMR technology provide spectral data that are highly amenable to analysis by computer software systems. Automated methods of analysis use constraint satisfaction, pseudoenergy minimization, directed search, neural net, simulated annealing, and/or genetic algorithms to establish sequential links and sequence-specific assignments. The most advanced systems provide automated analysis of complete backbone and extensive side-chain resonance assignments for proteins of 50–150 amino acids.