Katharina A. Zweig

Alexander Wilhelm, Katharina A. Zweig

In recent years, the number of new applications for highly complex AI systems has risen significantly. Algorithmic decision-making systems (ADMs) are one of such applications, where an AI system replaces the decision-making process of a human expert. As one approach to ensure fairness and transparency of such systems, explainable AI (XAI) has become more important. One variant to achieve explainability are surrogate models, i.e., the idea to train a new simpler machine learning model based on the input-output-relationship of a black box model. The simpler machine learning model could, for example, be a decision tree, which is thought to be intuitively understandable by humans. However, there is not much insight into how well the surrogate model approximates the black box. Our main assumption is that a good surrogate model approach should be able to bring such a discriminating behavior to the attention of humans; prior to our research we assumed that a surrogate decision tree would identify such a pattern on one of its first levels. However, in this article we show that even if the discriminated subgroup - while otherwise being the same in all categories - does not get a single positive decision from the black box ADM system, the corresponding question of group membership can be pushed down onto a level as low as wanted by the operator of the system. We then generalize this finding to pinpoint the exact level of the tree on which the discriminating question is asked and show that in a more realistic scenario, where discrimination only occurs to some fraction of the disadvantaged group, it is even more feasible to hide such discrimination. Our approach can be generalized easily to other surrogate models.

Tom Hanika, Marek Herde, Jochen Kuhn et al.

The field of collaborative interactive learning (CIL) aims at developing and investigating the technological foundations for a new generation of smart systems that support humans in their everyday life. While the concept of CIL has already been carved out in detail (including the fields of dedicated CIL and opportunistic CIL) and many research objectives have been stated, there is still the need to clarify some terms such as information, knowledge, and experience in the context of CIL and to differentiate CIL from recent and ongoing research in related fields such as active learning, collaborative learning, and others. Both aspects are addressed in this paper.

Sude Tavassoli, Katharina Anna Zweig

To understand a node's centrality in a multiplex network, its centrality values in all the layers of the network can be aggregated. This requires a normalization of the values, to allow their meaningful comparison and aggregation over networks with different sizes and orders. The concrete choices of such preprocessing steps like normalization and aggregation are almost never discussed in network analytic papers. In this paper, we show that even sticking to the most simple centrality index (the degree) but using different, classic choices of normalization and aggregation strategies, can turn a node from being among the most central to being among the least central. We present our results by using an aggregation operator which scales between different, classic aggregation strategies based on three multiplex networks. We also introduce a new visualization and characterization of a node's sensitivity to the choice of a normalization and aggregation strategy in multiplex networks. The observed high sensitivity of single nodes to the specific choice of aggregation and normalization strategies is of strong importance, especially for all kinds of intelligence-analytic software as it questions the interpretations of the findings.