Patrice Boizumault

Constraint programming for mining n-ary patterns

The aim of this paper is to model and mine patterns combining several local patterns (n-ary patterns). First, the user expresses his/her query under constraints involving n-ary patterns. Second, a constraint solver generates the correct and complete set of solutions. This approach enables to model in a flexible way sets of constraints combining several local patterns and it leads to discover patterns of higher level. Experiments show the feasibility and the interest of our approach.

Mining (Soft-) Skypatterns Using Dynamic CSP

Within the pattern mining area, skypatterns enable to express a user-preference point of view according to a dominance relation. In this paper, we deal with the introduction of softness in the skypattern mining problem. First, we show how softness can provide convenient patterns that would be missed otherwise. Then, thanks to Dynamic CSP, we propose a generic and efficient method to mine skypatterns as well as soft ones. Finally, we show the relevance and the effectiveness of our approach through a case study in chemoinformatics and experiments on UCI benchmarks.

A constraint language for declarative pattern discovery

Discovering pattern sets or global patterns is an attractive issue from the pattern mining community in order to provide useful information. By combining local patterns satisfying a joint meaning, this approach produces patterns of higher level and thus more useful for the end-user than the usual local patterns. In parallel, recent works investigating relationships between data mining and constraint programming (CP) show that the CP paradigm is a powerful framework to model and mine patterns in a declarative and generic way. We present a constraint-based language which enables us to define queries in a declarative way addressing patterns sets and global patterns. By specifying what the task is, rather than providing how the solution should be computed, it is easy to process by stepwise refinements to successfully discover global patterns. The usefulness of the approach is highlighted by several examples coming from the clustering based on associations. All primitive constraints of the language are modeled and solved using the SAT framework. We illustrate the efficiency of our approach through several experiments.

On-line resources allocation for ATM networks with rerouting

This paper presents an application we developed for France Telecom R&D to solve a difficult real-life network problem. The problem takes place in an Asynchronous Transfer Mode (ATM) network administration context and consists in planning demands of connection over a period of 1 year. A new demand is accepted if both bandwidth and Quality of Service (QoS) requirements are satisfied. Demands are not known prior to the assignment and must be performed on-line according to their arrival. Moreover, the acceptance or the reject of a demand must be decided within a given time of 1 min.First, we look for a route satisfying the new demand. In case of failure, we try to reroute some already accepted connections in order to satisfy this new demand. Rerouting has been modelled as a Weighted Constraint Satisfaction Problem (wcsp) and solved by VNS/LDS + CP, a hybrid method well suited for solving wcsps in on-line contexts. Experiments show that our rerouting enables to accept an average of 67% of demands that would be rejected otherwise.

PREFIX-PROJECTION Global Constraint for Sequential Pattern Mining

Sequential pattern mining under constraints is a challenging data mining task. Many efficient ad hoc methods have been developed for mining sequential patterns, but they are all suffering from a lack of genericity. Recent works have investigated Constraint Programming CP methods, but they are not still effective because of their encoding. In this paper, we propose a global constraint based on the projected databases principle which remedies to this drawback. Experiments show that our approach clearly outperforms CP approaches and competes well with ad hoc methods on large datasets.

Mining Relevant Sequence Patterns with CP-Based Framework

Sequential pattern mining under various constraints is a challenging data mining task. The paper provides a generic framework based on constraint programming to discover sequence patterns defined by constraints on local patterns (e.g., Gap, regular expressions) or constraints on patterns involving combination of local patterns such as relevant subgroups and top-k patterns. This framework enables the user to mine in a declarative way both kinds of patterns. The solving step is done by exploiting the machinery of Constraint Programming. For complex patterns involving combination of local patterns, we improve the mining step by using dynamic CSP. Finally, we present two case studies in biomedical information extraction and stylistic analysis in linguistics.

A Value Ordering Heuristic for Weighted CSP

Cellular learning automata (CLA) which is obtained by combining cellular automata (CA) and learning automata (LA) models is a mathematical model for dynamical complex systems that consists of a large number of simple learning components. CLA- EC, introduced recently is an evolutionary algorithm which is obtained by combining CLA and evolutionary computation (EC). In this paper CLA-EC with recombination operator is introduced. Recombination increases explorative behavior of CLA-EC and also provides a mechanism for partial structure exchange between chromosomes of population individuals that standard CLA-EC is not capable of performing it. This modification greatly improves CLA-EC ability to effectively search solution space and leave local optima. Experimental results on five optimization test functions show the superiority of this new version of CLA-EC over the standard CLA-EC.

Constrained clustering using SAT

Constrained clustering - finding clusters that satisfy user-specified constraints - aims at providing more relevant clusters by adding constraints enforcing required properties. Leveraging the recent progress in declarative and constraint-based pattern mining, we propose an effective constraint-clustering approach handling a large set of constraints which are described by a generic constraint-based language. Starting from an initial solution, queries can easily be refined in order to focus on more interesting clustering solutions. We show how each constraint (and query) is encoded in SAT and solved by taking benefit from several features of SAT solvers. Experiments performed using MiniSat on several datasets from the UCI repository show the feasibility and the advantages of our approach.

Soft Threshold Constraints for Pattern Mining

Constraint-based pattern discovery is at the core of numerous data mining tasks. Patterns are extracted with respect to a given set of constraints (frequency, closedness, size, etc). In practice, many constraints require threshold values whose choice is often arbitrary. This difficulty is even harder when several thresholds are required and have to be combined. Moreover, patterns barely missing a threshold will not be extracted even if they may be relevant. In this paper, by using Constraint Programming we propose a method to integrate soft threshold constraints into the pattern discovery process. We show the relevance and the efficiency of our approach through a case study in chemoinformatics for discovering toxicophores.

Guiding VNS with Tree Decomposition

Tree decomposition introduced by Robertson and Seymour aims to decompose a problem into clusters constituting an a cyclic graph. There are works exploiting tree decomposition for complete search methods. In this paper, we show how tree decomposition can be used to efficiently guide the exploration of local search methods that use large neighborhoods like VNS. We introduce tightness dependent tree decomposition which allows to take advantage of both the structure of the problem and the constraints tightness. Experiments performed on random instances (GRAPH) and real life instances (CELAR and SPOT5) show the appropriateness and the efficiency of our approach.