From Black and White to Full Colour: Extending Redescription Mining Outside the Boolean World

Abstract

Redescription mining is a powerful data analysis tool that is used to find multiple descriptions of the same entities. Consider geographical regions as an example. They can be characterized by the fauna that inhabits them on one hand and by their meteorological conditions on the other hand. Finding such redescriptors, a task known as niche-finding, is of much importance in biology.But current redescription mining methods cannot handle other than Boolean data. This restricts the range of possible applications or makes discretization a prerequisite, entailing a possibly harmful loss of information. In nichefinding, while the fauna can be naturally represented using a Boolean presence/absence data, the weather cannot.In this paper, we extend redescription mining to realvalued data using a surprisingly simple and efficient approach. We provide extensive experimental evaluation to study the behaviour of the proposed algorithm. Furthermore, we show the statistical significance of our results using recent innovations on randomization methods.

Finding relational redescriptions

“…Relational pattern languages are thus a natural candidate for redescriptions, but existing approaches have focused on propositional features (Ramakrishnan et al, 2004) and real-valued attributes (Galbrun and Miettinen, 2012). They operate on matrices with a row for every object in the data and a column for every attribute or feature.…”

“…These approaches have been shown to find simple redescriptions of single nodes in a bibliographic network, describing a researcher either in terms of the conferences he publishes at, or in terms of his co-authors, that is, using only attributes in the form of the labels of neighboring nodes (Gallo et al, 2008;Galbrun and Miettinen, 2012). However, as we illustrate in Section 6, considering pairs of objects and features based on their connections in terms of complex longer distance relations would inflate the size of the matrix and thus the search space of the algorithm.…”

“…We next compare to an existing propositional redescription mining algorithm, following ideas from propositionalization (Kuzelka and Zelezný, 2009;Dinh et al, 2012) to suitably transform the data. Specifically, we extracted features from the dataset by enumerating paths up to a given length joining each pair of objects and ran the ReReMi propositional redescription mining algorithm (Galbrun and Miettinen, 2012) on the resulting propositional dataset. We considered paths of length one and two for all datasets.…”

Towards Finding Relational Redescriptions

“…This clause has query variables #A and #Z and is linked due to the spouse and parent edges. 3 Its attribute set is {sex, gen, age}.…”

“…One way of creating such links is to find sets of objects together with their descriptions in different terminologies, as done in redescription mining [2,3]. However, so far, this technique has only considered propositional or real valued attributes.…”