Céline Fiot

Extended Time Constraints for Sequence Mining

Many applications require techniques for temporal knowledge discovery. Some of those approaches can handle time constraints between events. In particular some work has been done to mine generalized sequential patterns. However, such constraints are often too crisp or need a very precise assessment to avoid erroneous information. Therefore, in this paper we propose to soften temporal constraints used for generalized sequential pattern mining. To handle these constraints while data mining, we design an algorithm based on sequence graphs. Moreover, as these relaxed constraints may extract more generalized patterns, we propose temporal accuracy measure for helping the analysis of the numerous discovered patterns.

Softening the blow of frequent sequence analysis: soft constraints and temporal accuracy

Mining temporal knowledge has many applications. Such knowledge can be all the more interesting as some time constraints between events can be integrated during the mining task. Both in data mining and machine learning, some methods have been proposed to extract and manage such knowledge using temporal constraints. In particular, some work has been done to mine Generalised Sequential Patterns (GSPs). However, such constraints are often too crisp or need a very precise assessment to avoid erroneous information. Within this context, we propose an approach based on sequence graphs derived from soft temporal constraints. These relaxed constraints enable us to find more GSPs. We also propose a temporal accuracy measure to provide the user with a tool for analysing the numerous extracted patterns.

Web Access Log Mining with Soft Sequential Patterns

Mining the time-stamped numerical data contained in web access logs is interesting for numerous applications (e.g. customer targeting, automatic updating of commercial websites or web server dimensioning). In this context, the algorithms for sequential patterns mining do not allow processing numerical information frequently. In previous works we defined fuzzy sequential patterns to cope with the numerical representation problem. In this paper, we apply these algorithms to web mining and assess them through experiments showing the relevancy of this work in the context of web access log mining.

Learning Bayesian network structure from incomplete data without any assumption

Since most real-life data contain missing values, reasoning and learning with incomplete data has become crucial in data mining and machine learning. In particular, Bayesian networks are one machine learning technique that allows for reasoning with incomplete data, but training such networks on incomplete data may be a difficult task. Many methods were thus proposed to learn Bayesian network structure from incomplete data, based on multiple structure generation and scoring of their adequacy to the dataset. However, this kind of approaches may be time-consuming. Therefore we propose an efficient dependency analysis approach that uses a redefinition of probability calculation to take incomplete records into account while learning BN structure, without generating multiple possibilities. Some experiments on well-known benchmarks are described to show the validity of our proposal.