David Lesaint

iOpt: A Software Toolkit for Heuristic Search Methods

Heuristic Search techniques are known for their efficiency and effectiveness in solving NP-Hard problems. However, there has been limited success so far in constructing a software toolkit which is dedicated to these methods and can fully support all the stages and aspects of researching and developing a system based on these techniques. Some of the reasons for that include the lack of problem modelling facilities and domain specific frameworks which specifically suit the operations of heuristic search, tedious code optimisations which are often required to achieve efficient implementations of these methods, and the large number of available algorithms - both local search and population-based - which make it difficult to implement and evaluate a range of techniques to find the most efficient one for the problem at hand. The iOpt Toolkit, presented in this article, attempts to address these issues by providing problem modelling facilities well-matched to heuristic search operations, a generic framework for developing scheduling applications, and a logically structured heuristic search framework allowing the synthesis and evaluation of a variety of algorithms. In addition to these, the toolkit incorporates interactive graphical components for the visualisation of problem and scheduling models, and also for monitoring the run-time behaviour and configuring the parameters of heuristic search algorithms.

Solving a Telecommunications Feature Subscription Configuration Problem

Call control features (e.g., call-divert, voice-mail) are primitive options to which users can subscribe off-line to personalise their service. The configuration of a feature subscription involves choosing and sequencing features from a catalogue and is subject to constraints that prevent undesirable feature interactions at run-time. When the subscription requested by a user is inconsistent, one problem is to find an optimal relaxation. In this paper, we show that this problem is NP-hard and we present a constraint programming formulation using the variable weighted constraint satisfaction problem framework. We also present simple formulations using partial weighted maximum satisfiability and integer linear programming. We experimentally compare our formulations of the different approaches; the results suggest that our constraint programming approach is the best of the three overall.

Consistency Techniques for Finding an Optimal Relaxation of a Feature Subscription

Telecommunication services are playing an increasing and potentially disruptive role in our lives. As a result, service providers seek to develop personalisation solutions that put customers in charge of controlling and enriching their services. In this context, the personalisation approach consists of exposing a catalogue of call control features (e.g., call-divert, voice-mail) to end-users and letting them subscribe to a subset of features subject to a set of precedence and exclusion constraints. When a subscription is inconsistent, the problem is to find an optimal relaxation. We present a constraint programming formulation to find an optimal reconfiguration of features. We investigate the performance of maintaining arc consistency within branch and bound search. We also study the impact of maintaining mixed consistency, that is maintaining different levels of consistency on different sets of variables. We further present a global constraint and a set of filtering rules that exploit the structure of our problem. We theoretically and experimentally compare all approaches. Our results demonstrate that the filtering rules of the global constraint outperform all other approaches when a catalogue is dense, and mixed consistency pays off when a catalogue is sparse.

Context-sensitive call control using constraints and rules

Personalisation and context-awareness are fundamental concerns in Telephony. This paper introduces a rule-based system - 4CRULES - which enables context-sensitive call control by the means of feature configuration rules. 4CRULES is interoperable with standard context services and compositional feature architectures. It has been designed to resolve feature interactions, manage conflicting preferences, and mitigate the uncertainty affecting context data. This is achieved through a constraint optimisation model that maximises adherence to user requirements and domain constraints. Experiments on a suite of instances confirm the practicality of the approach and highlight performance- and adherence-critical factors.

Aspects for Synthesizing Applications by Refinement

GenVoca is a powerful model for component-based product-line architectures (PLAs) advocating stepwise refinement as a composition principle. This paper introduces a refinement-oriented generative language – ReGaL – to implement statically configurable GenVoca PLAs. Whereas components are programmed in Java, refinements are program-med in ReGaL by the means of generic aspects. Applications are themselves specified in ReGaL as type equations. ReGaL compiles type equations by instantiating and weaving refinement aspects with components to synthesize the requested (Java) application. As opposed to template-based generative implementations, ReGaL promotes a clean separation of components and refinements, hence eliminating code tangling and related issues. It also defers the choice of component class composition structures until configuration time, which provides added flexibility to adapt applications. Besides, its architecture model enforces a clear role-based design of components and supports useful architectural patterns.Step-wise refinement is a powerful paradigm for developing a complex program from a simple program by adding features incrementally. We present the AHEAD (Algebraic Hierarchical Equations for Application Design) model that shows how step-wise refinement scales to synthesize multiple programs and multiple non-code representations. AHEAD shows that software can have an elegant, hierarchical mathematical structure that is expressible as nested sets of equations. We review a tool set that supports AHEAD. As a demonstration of its viability, we have bootstrapped AHEAD tools solely from equational specifications, generating Java and non-Java artifacts automatically, a task that was accomplished only by ad hoc means previously.

Dynamic Planner: A Decision Support Tool for Resource Planning

Accurate resource planning is critical to a company’s performance and profitability. This paper describes the motivation for and the realisation of an automated resource planning system (i.e. Dynamic Planner) within British Telecom. Dynamic Planner is a decision support tool to help resource managers decide how best to deploy their resources. It provides a feel for what resources are needed; how many are needed and how best they should be deployed. It also provides the framework to compare the impact of scenarios involving overtime, borrowed resources and skill move. Dynamic Planner is built atop iOpt, a Java-based optimisation toolkit for modelling and solving combinatorial problems using invariants (one-way constraints) and heuristic search methods respectively.

Model and Combinatorial Optimization Methods for Tactical Planning in Closed-Loop Supply Chains

Distribution planning in closed-loop supply chains is concerned with determining transfer and repair operations based on demand forecasts and subject to backordering, inventory, transfer and repair constraints. We present a mixed-integer programming model and a dedicated metaheuristics for this problem and show it is is NP-hard. The model is applicable to a wide range of closed-loop supply chains with different network topologies and site functions and it can also support different planning strategies by means of a weighted objective function. Comparative experiments on pseudo-random instances built on a case study in telecommunication service operations demonstrate the effectiveness and scalability of the metaheuristics. Lastly, we discuss possible extensions to address common supply chain requirements, including the ability to produce robust plans in uncertain environments.

A bottom-up implementation of Path-Relinking for Phylogenetic reconstruction applied to Maximum Parsimony

In this article we describe a bottom-up implementation of Path-Relinking for Phylogenetic Trees in the context of the resolution of the Maximum Parsimony problem with Fitch optimality criterion. This bottom-up implementation is compared to two versions of an existing top-down implementation. We show that our implementation is more efficient, more interesting to compare trees and to give an estimation of the distance between two trees in terms of the number of transformations.

A Decomposition Approach for Discovering Discriminative Motifs in a Sequence Database

Considerable effort has been invested over the years in ad-hoc algorithms for item set and pattern mining. Constraint programming has recently been proposed as a means to tackle item set mining tasks within a general modelling framework. We follow this approach to address the discovery of discriminative n-ary motifs in databases of labeled sequences. We define a n-ary motif as a mapping of n patterns to n class-wide embeddings and we restrict the interpretation of constraints on a motif to the sequences embedding all patterns. We formulate core constraints that minimize redundancy between motifs and introduce a general constraint optimization framework to compute common and exclusive motifs. We cast the discovery of closed and replication-free motifs in this framework for which we propose a two-stage approach based on constraint programming. Experimental results on datasets of protein sequences demonstrate the efficiency of the approach.

A decomposition approach for discovering discriminative motifs in a sequence database

Considerable effort has been invested over the years in ad-hoc algorithms for item set and pattern mining. Constraint programming has recently been proposed as a means to tackle item set mining tasks within a general modelling framework. We follow this approach to address the discovery of discriminative n-ary motifs in databases of labeled sequences. We define a n-ary motif as a mapping of n patterns to n class-wide embeddings and we restrict the interpretation of constraints on a motif to the sequences embedding all patterns. We formulate core constraints that minimize redundancy between motifs and introduce a general constraint optimization framework to compute common and exclusive motifs. We cast the discovery of closed and replication-free motifs in this framework for which we propose a two-stage approach based on constraint programming. Experimental results on datasets of protein sequences demonstrate the efficiency of the approach.