27 May 2019. Artificial Intelligence (AI) has allowed an unprecedented level of automation in our society, powering computers to execute with reasonable confidence tasks previously restricted to humans, such as translating a text or recognising objects in a photograph. AI has also been crucial for identifying patterns in very large data sets in tasks out of the human capabilities, like fraud detection and recommender systems.

However, in no small extent, this high performance comes with the cost of transparency. AI-complex statistical models and neural networks are generated from an intricate set of computational operations from which there is no easy way to associate a given input data and its impact in the trained model. For this reason, AI has been frequently called a “black-box” technology.

Rules to prevent bias and unfair behaviour
Afraid of the impact of those opaque models, the European Parliament defined legal rules to prevent bias and unfair behaviour. Also, this fear is not without reasons. Many studies have shown that, as AI models learn patterns from human-generated data sets, they are also learning the bias present in society. As a simple example, the de-facto method to represent words as vectors, called word embeddings, creates a higher association between the words “man” and “success”, than “woman” and “success”. Other study has also shown that having a name frequently used by a minority group, such as Ebony for Afro-Americans, has a negative impact in AI-based sentiment analysis, compared to a name associated with European Americans, like Amanda. Other types of prejudice were also found in credit-score systems when assessing loan risk.

Explaining machine-learning models
Given these challenges, in recent years, the research area devoted to explaining machine learning models gained strength in the AI field. The main idea is to provide meaningful information about the algorithms in such a manner that humans without a technical background can understand how it works and potentially identify unfairness and bias. The area of Explainable AI, moreover, goes even further.

Zachary C. Lipton defined a comprehensive taxonomy of explanations in the context of AI, highlighting various criteria of classification such as motivation (trust, causality, transferability, informativeness and fairness & ethics) and property (transparency and post-hoc interpretability).

Trust is, by far the most common motivation presented in the literature. Or Biran and Courtenay Cotton, for instance, show that users demonstrate higher confidence when using a system that they understand how it works. Fairness & Ethics is also a strong driver as the well-known European General Data Protection Regulation (GDPR), which guarantees both rights “for meaningful information about the logic involved” and “to non-discrimination” to prevent bias and unfair behaviour, mainly targeting decision-making algorithms. Although less representative, explanations are also used to support users’ feedbacks to intelligent systems.
From the property criterion, transparency allows understanding the algorithm’s mechanism of decision, by contemplating “the entire model at once” and understanding each of its parts and its learning mechanism. Typical methods complying with these requirements are the so-called “explainable by design” such as linear regression, decision trees and rule-based approaches when dealing with small models.

Post-hoc explanations, on the other hand, make use of interpretations to deliver meaningful information about the AI model. Instead of showing how the model works, it presents evidence of its rationale by making use of (i) textual descriptions, (ii) visualisations able to highlight parts of an image from which the decision was made, (iii) 2D-representation of high-dimensional spaces or (iv) explanation by similarity. Although this type of explanation does not tell precisely how the output was generated, it still presents useful information about its internal mechanism.

Post-hoc explanations of AI models
One of the works developed in our research team offers an example of a post-hoc explanation for the task of text entailment, in which, based on a given fact the system evaluates whether a second statement (the hypothesis) is true or false. In this work, the approach makes use of a graph representation of the words’ meanings to navigate between the definitions based on a semantic similarity measure.

For instance, assuming the fact “IBM cleared $18.2 billion in the first quarter”, and the hypothesis “IBM’s revenue in the first quarter was $18.2 billion”, it provides not only a yes/no answer but also an explanation: Yes, it entails because “To clear is to yield as a net profit” and “Net profit is synonym of revenue”.

Improving text entailment systems affects the performance of many natural language processing tasks such as question answering, text summarization, and information extraction, among others. Whereas this task seems trivial for a human, in fact, it represents the essence of the challenge that modern artificial intelligence models face: understanding meaning.

Siegfried Handschuh is full Professor of Data Science at the University of St.Gallen.

Image: Adobe Stock/Oleksii

Prof. Dr. Siegfried Handschuh, Spezialist für Data Science an der Universität St.Gallen entwickelte gemeinsam mit einem interdisziplinären Team, bestehend aus dem Historiker Simon Donig und den IT-Spezialisten Maria Christoforaki und Bernhard Bermeitinger, einen Algorithmus, der klassizistische Möbel und Kunstwerke identifiziert und kategorisiert. Die Künstliche Intelligenz lernte kontinuierlich, weitere Gegenstände einzuordnen. Damit können künftig grosse Datenmengen in den Geisteswissenschaften klassifiziert werden und Merkmale einer bestimmten Epoche korrigiert bzw. gefestigt werden. Wie das konkret gelingt, erklärt Siegfried Handschuh im Video-Interview. Dieses Video erschien zuerst in der HSG-Focus-Ausgabe «Kunst».

Lesen Sie hier die gesamte Ausgabe: http://magazin.hsgfocus.ch
© Universität St.Gallen (HSG)

26 October 2018. Moderator Jacqueline Gasser-Beck, Head of the Teaching Innovation Lab, reminded the audience at the outset of the panel discussion that the St. Gallen-Appenzell Chamber of Industry and Commerce (IHK) had commissioned the HSG to draw up a feasibility study to determine whether the introduction of a new major would make sense.

The conclusion of the study was unequivocal
According to Walter Brenner, Full Professor of Information Management, the result of the study was unequivocal. “Digitalisation will have a great impact on society and the economy. Therefore there is no getting around IT and software knowledge,” he emphasised at the panel discussion and described the IT major as a great opportunity for the University of St.Gallen. The HSG would thus fulfil its obligation to provide students with the necessary toolkit for everyday professional life while ensuring that the HSG would not lose its appeal in an age of increasing digitalisation.

Simon Mayer, one of the three new IT professors, explained the planned course of action. On the one hand, all students should be obliged to acquire a fundamental knowledge of business IT. On the other hand, the establishment of an IT major with an undergraduate and a Master’s programme is in the pipeline. This, however, will be contingent on the IT Education Initiative being approved by the voting public on the occasion of the cantonal ballot in early 2019. In order to save time, autumn 2019 will see the start of the Master’s programme first, with the undergraduate programme following later.

Combining business and IT
Education was the central leverage factor of economic policy in Eastern Switzerland, said Frank Bodmer, Head of IHK Research, which was why it was a great concern of the Chamber of Industry to counteract the shortage of specialists in the IT sector. He was pleased by the fast response to this concern and by how quickly work was taken up. He considered the planned approach to be practice-oriented and sensible.

IT specialist Michèle Mégroz, President of the HSG Alumni Eastern Switzerland Chapter and CEO of CSP AG, also praised the approach. The combination of business and IT was gaining ever greater significance. There were hardly any projects any longer which did not also require IT knowledge. Conversely, it was very difficult to find people with in-depth knowledge of both fields. She was convinced that the University of St.Gallen would profit from a new major in IT.

Student Sebastian Kuhn also welcomed the approach. To date, students who are interested in IT had tried to acquire IT knowledge outside their studies. He felt that the debate about the IT Education Offensive was creating an optimistic mood.

Three IT professors appointed
The initial steps towards the introduction of IT into the HSG’s degree courses were made with the appointment of three IT professors. They took up their posts in late summer. On the Digital Day, they presented the focal points of their teaching and research activities. Siegfried Handschuh is Full Professor of Data Science. His research focuses on a combination of compositional linguistic methods and machine learning techniques. At the University of St.Gallen, he will also deal with the language problem of artificial intelligence and the idea of the automated data scientist. In the area of applied research, he aimed to cooperate with the core disciplines of the University, explained Siegfried Handschuh.

Simon Mayer concentrates his research on the industrial internet of things. His work ranges from issues of the web-based interaction of different cyber-physical systems among each other and with people. Applications in this field can be found in the haulage industry, but primarily also in industrial production and the issues of Industry 4.0. The focus of the employment of the Full Professor of Interaction- and Communication-Based Systems is on the establishment of a business IT department, with an emphasis on teaching and research in the fields of distributed systems, the internet of things and human-machine interaction.

The third IT professor is called Damian Borth and is Full Professor of Artificial Intelligence and Machine Learning. The focus of his employment is on the establishment of the new IT Department and major in IT. He is regarded as one of Europe’s most successful young academics in the field of artificial intelligence. One focus of his work is the analysis of large volumes of unstructured data such as texts, images, videos or time series with the help of deep neural networks. IT had rather a bad image in the German-speaking area, said Damian Borth. This image must urgently be changed in order to counteract the shortage of specialists.

Photo: Fotolia/ Siarhei