Abstract
The information contained in companies' financial statements is valuable to several users. Much of the relevant information in such documents is contained in tables and is currently mainly extracted by hand. We propose a method that accomplishes a prior step of the task of automatically extracting information from tables in documents: selecting the lines that are likely to belong to tables. Our method has been developed by empirically analyzing a set of Portuguese companies' financial statements using statistical and data mining techniques. Empirical evaluation indicates that more than 99% of table lines are selected after discarding at least 50% of all lines. The method can cope with the complexity of styles used in assembling information on paper and adapt its performance accordingly, thus maximizing its results.

Abstract
The emerging standard for the platform- and system-independent representation of data mining models PMML (Predictive Model Markup Language) is currently supported by a number of knowledge discovery support engines. The primary purpose of the PMML standard is to separate model generation from model storage in order to enable users to view, post-process, and utilize data mining models independently of the tool that generated the model. In this paper two systems, called VizWiz and PEAR, are described. These software packages allow for the visualization and evaluation of data mining models that are specified in PMML. They can be viewed. as decision support systems, since they enable non-expert users of data mining results to interactively inspect and evaluate these results.

Abstract
This paper describes a method designed for data mining applications where the main goal is to predict extreme and rare values of a continuous target variable, as well as to understand under which conditions these values occur. Our objective is to induce models that are accurate at predicting these outliers but are also interpretable from the user perspective. We describe a new splitting criterion for regression trees that enables the induction of trees achieving these goals. We evaluate our proposal on several real world problems and contrast the obtained models with standard regression trees. The results of this evaluation show the clear advantage of our proposal in terms of the evaluation statistics that are relevant for these applications.

Abstract
In several applications the main interest resides in predicting rare and extreme values. This is the case of the prediction of harmful algae blooms. Though it's rare, the occurrence of these blooms has a strong impact in river life forms and water quality and turns out to be a serious ecological problem. In this paper, we describe a data mining method whose main goal is to predict accurately this kind of rare extreme values. We propose a new splitting criterion for regression trees that enables the induction of trees achieving these goals. We carry out an analysis of the results obtained with our method on this application domain and compare them to those obtained with standard regression trees. We conclude that this new method achieves better results in terms of the evaluation statistics that are relevant for this kind of applications.

Abstract

Abstract