Italy

The interest in Explainable Artificial Intelligence (XAI) research is dramatically grown during the last few years. The main reason is the need of having systems that beyond being effective are also able to describe how a certain output has been obtained and to present such a description in a comprehensive manner with respect to the target users. A promising research direction making black boxes more transparent is the exploitation of semantic information. Such information can be exploited from different perspectives in order to provide a more comprehensive and interpretable representation of AI models. In this paper, we focus on one of the key components of the semantic-based explanation generation process: the explanation graph. We discuss its role and how it can work has a bridge for making an explanation more understandable by the target users, complete, and privacy preserving. We show how the explanation graph can be integrated into a real-world solution and we discuss challenges and future work.

AIxIA 2021

Bridging Signals to Natural Language Explanations With Explanation Graphs

workshop paper

[![](https://assets.underline.io/uploads/markdown_image/1/image/9a88ae95fbde06e80f4bcf6d4f7d133a.png)](https://certificateaixia.paperform.co/)
The certificate will be provided by December 17. 
In case you need it urgently please contact: francesca.gasparini@unimib.it. 

 

The 20th International Conference of the Italian Association for Artificial Intelligence (**AIxIA 2021**) is organized by AIxIA (Associazione Italiana per l’Intelligenza Artificiale), which is a non-profit scientific society founded in 1988 and devoted to the promotion of Artificial Intelligence. The society aims to increase the public awareness of AI, encourage the teaching of it and promote research in the field.

 
 



**General Chair**

Stefania Bandini, University of Milano-Bicocca, Italyand RCAST - The University of Tokyo, Japan


**Program Chairs**

Viviana Mascardi, University of Genova, Italy; Matteo Palmonari, University of Milano-Bicocca, Italy; Giuseppe Vizzari, University of Milano-Bicocca, Italy

**Organizing Committee**

Francesca Gasparini, University of Milano-Bicocca, Italy (**Chair**); Stefano Ferilli, University of Bari, Italy; Marta Giltri, University of Milano-Bicocca, Italy; Daniela Briola, University of Insubria, Italy


**Publicity and Communication Committee**

Elisabetta Fersini, University of Milano-Bicocca, Italy (**Chair**); Angelo Ferrando, University of Genova, Italy; Francesca Alessandra Lisi, University of Bari, Italy; Giovanni Stilo, University of L'Aquila, Italy; Marco Viviani, University of Milano-Bicocca, Italy


**Workshop Chairs**

Gabriella Cortellessa, ISTC-CNR, Rome, Italy; Silvia Rossi, University of Naples Federico II, Italy; Fabio Stella, University of Milano-Bicocca, Italy


**PC Members**

Fulvio Mastrogiovanni, University of Genova, Italy; Roberto Basili, University of Rome Tor Vergata, Italy; Nicola Gatti, Polytechnic University of Milan, Italy; Stefano Bistarelli, University of Perugia, Italy; Piero Poccianti, Consorzio Operativo Gruppo MPS, Italy; Francesco Scarcello, University of Calabria, Italy; Flavio S. Correa Da Silva, University of São Paulo, Brazil; Elena Cabrio, Université Côte d'Azur, CNRS, Inria, I3S, France; Antonio Lieto, University of Turin, Italy; Diego Calvanese, Free University of Bozen-Bolzano, Italy; Alessio Micheli, University of Pisa, Italy; Francesca A. Lisi University of Bari Aldo Moro, Italy; Mauro Vallati, University of Huddersfield, United Kingdom; Bernardo Magnini, Fondazione Bruno Kessler (FBK), Italy; Luigia C. Aiello, Sapienza University of Rome, Italy; Rafael H. Bordini, Pontifical Catholic University of Rio Grande do Sul, Brazil; Chiara Ghidini, Fondazione Bruno Kessler (FBK), Italy; Federico Bergenti, University of Parma, Italy; Francesco Ricca, University of Calabria, Italy; Andrea Burattin, Technical University of Denmark, Denmark; Angelo Oddi, ISTC-CNR, Italian National Research Council, Italy; Farshad Badie, Aalborg University, Denmark; Emanuel Sallinger, University of Oxford, United Kingdom; Marin Lujak, University Rey Juan Carlos, Spain; Giancarlo Fortino, University of Calabria, Italy; Salvatore Gaglio, University of Palermo, Italy; Johannes K. Fichte, TU Dresden, Germany; Matteo Baldoni, University of Turin, Italy; Fabrizio Riguzzi, Univeristy of Ferrara, Italy; Ivana Dusparic, Trinity College Dublin, Ireland; Serena Villata, CNRS - Laboratoire d'Informatique, Signaux et Systèmes de Sophia-Antipolis, France; Paolo Rosso, Technical University of Valencia, Spain; Ana L. C. Bazzan, Federal University of Rio Grande do Sul, Brazil; Gianluigi Greco, University of Calabria, Italy; Stefano Cagnoni, University of Parma, Italy; Gian Luca Pozzato, University of Turin, Italy; Salvatore Ruggieri, University of Turin, Italy; Claudio Di Ciccio, Sapienza University of Rome, Italy; Eloisa Vargiu, Cetaqua, Water Technology Center, Spain; Marco Maratea, University of Genova, Italy; Andrea Roli, University of Bologna, Italy; Viviana Patti, University of Turin, Italy; Stefania Costantini, University of L'Aquila, Italy; Antonius Weinzierl, Vienna University of Technology, Austria; Silvia Rossi, University of Naples Federico II, Italy; Giovanni Semeraro, University of Bari, Italy; Luciano Serafini, Fondazione Bruno Kessler (FBK), Italy; Laura Giordano, University of Eastern Piedmont, Italy; Ernesto Jimenez-Ruiz, City, University of London, United Kingdom; Marco Gavanelli, University of Ferrara, Italy; Federico Chesani, University of Bologna, Italy

Please register with AIxIA to be able to enter the conference on Underline.io. If you have already registered, please login to Underline.io to be able to see full conference content. In case you do not have an Underline account, please create one. Thank you

Please login to your Underline account

The 20th International Conference of the Italian Association for Artificial Intelligence (AIxIA 2021) is organized by AIxIA (Associazione Italiana per l’Intelligenza Artificiale), which is a non-profit scientific society founded in 1988 and devoted to the promotion of Artificial Intelligence. The society aims to increase the public awareness of AI, encourage the teaching of it and promote research in the field.

The rapidly growing research area of eXplainable Artificial Intelligence (XAI) focuses on making Machine Learning systems’ decisions more transparent and humanly understandable. One of the most successful XAI strategies is to provide explanations in terms of visualisations and, more specifically, low-level input features such as relevance scores or heat maps of the input, like sensitivity analysis or layer-wise relevance propagation methods. The main problem with such methods is that starting from the relevance of low-level features, the human user needs to identify the overall input properties that are salient. Thus, a current line of XAI research attempts to alleviate this weakness of low-level approaches, constructing explanations in terms of input features that represent more salient and understandable input properties for a user, which we call here Middle-Level input Features (MLF). In addition, another interesting and very recent approach is that of considering hierarchically organised explanations. Thus, in this paper, we investigate the possibility to combine both MLFs and hierarchical organisations. The potential advantages of providing explanations in terms of hierarchically organised MLFs are grounded on the possibility of exhibiting explanations to a different granularity of MLFs interacting with each other. We experimentally tested our approach on 300 Birds Species and Cars dataset. The results seem encouraging.

Explanations in terms of Hierarchically organised Middle Level Features

We tackle sequential learning under label noise in applications where a human supervisor can be queried to relabel suspicious examples. Existing approaches are flawed, in that they only relabel incoming examples that look "suspicious" to the model. As a consequence, those mislabeled examples that elude (or don't undergo) this cleaning step end up tainting the training data and the model with no further chance of being cleaned. We propose Cincer, a novel approach that cleans both new and past data by identifying pairs of mutually incompatible examples. Whenever it detects a suspicious example, Cincer identifies a counter-example in the training set that -- according to the model -- is maximally incompatible with the suspicious example, and asks the annotator to relabel either or both examples, resolving this possible inconsistency. The counter-examples are chosen to be maximally incompatible, so to serve as explanations of the model's suspicion, and highly influential, so to convey as much information as possible if relabeled. Cincer achieves this by leveraging an efficient and robust approximation of influence functions based on the Fisher information matrix (FIM). Our extensive empirical evaluation shows that clarifying the reasons behind the model's suspicions by cleaning the counter-examples helps in acquiring substantially better data and models, especially when paired with our FIM approximation.

Interactive Label Cleaning with Example-based Explanations

In this paper I will tackle the debated question of whether Deep Neural Networks “understand” Natural Language when they process it either for classification or for generation tasks. I will start by quoting work currently carried out in the field of explainable AI then I will look into the internal procedures adopted by current DNNs to cope with the enormous and untreatable dimension of vocabulary size due to the presence of rare words and the use of subwords or n-gram character sequences. To explain the inability of DNNs to reuse previously discovered knowledge I will deal with the issue of generalization and the lack of systematicity in learning models. I will also propose reasons to motivate the difficulty of generalizing with morphology-rich and weakly configurational languages like Italian. In particular vocabulary size and rare words need a different approach from the one used with English texts. Eventually I will propose some alternative ways to deal with the problems discussed in the paper

What's Wrong with Deep Learning for Meaning Understanding

Shapley values has established itself as one of the most appropriate and theoretically sound frameworks for explaining predictions from complex machine learning models. The popularity of Shapley values in the explanation setting is probably due to Shapley values’ unique theoretical properties. The main drawback with Shapley values, however, is that the computational complexity grows exponentially in the number of input features, making it unfeasible in many real world situations where there could be hundreds or thousands of features. Furthermore, with many (dependent) features, presenting/visualizing and interpreting the computed Shapley values also become challenging. The present paper introduces and showcases a method that we call groupShapley. The idea of the method is to group features and then compute and present Shapley values for these groups instead of for all individual features. Reducing hundreds or thousands of features to half a dozen or so feature groups makes precise computations practically feasible, and the presentation and knowledge extraction greatly simplified. We give practical advice for using the approach and illustrate its usability in three different real world examples. The examples vary in both data type (regular tabular data and time series), feature dimension (medium to high), and application (insurance, genetics, and banking).

Efficient and simple prediction explanations with groupShapley: A practical perspective

In recent years, there have been many attempts to combine XAI with the field of symbolic AI in order to generate explanations for neural networks that are more interpretable and better align with human reasoning, with one prominent candidate for this synergy being the sub-field of computational argumentation. One method is to represent neural networks with quantitative bipolar argumentation frameworks (QBAFs) equipped with a particular semantics. The resulting QBAF can then be viewed as an explanation for the associated neural network. In this paper, we explore a novel LRP-based semantics under a new QBAF variant, namely neural QBAFs (nQBAFs). Since an nQBAF of a neural network is typically large, the nQBAF must be simplified before being used as an explanation. Our empirical evaluation indicates that the manner of this simplification is all important for the quality of the resulting explanation.

LRP-Based Argumentative Explanations for Neural Networks

On the Applications of Evolutionary Algorithms to the Cybersecurity Domain

Digital images made a big difference in everyday life and science. They became crucial part in various areas such as medicine, astronomy, agriculture, etc. A common problem in applications with digital images is classification. Since recently, convolutional neural network has been widely used for image classification and classification accuracy is dramatically improved. To have the best possible classification accuracy of a CNN for a specific application, the necessary step is tuning various hyperparameters such as the number of different layers, number of neurons in each layer, optimization algorithm, activation functions, kernel size, optimization algorithm, etc. Recent studies show that swarm intelligence algorithms used for CNN’s hyperparameters tuning give promising results and further research is needed. In this paper, a short analysis of swarm intelligence algorithms and their application to tuning CNN hyperparameters is presented.

Swarm Intelligence Algorithms for Convolutional Neural Networks

Building the Building Blocks - Evolving Hyper-Features for Model Robustness and Readability

Closing

The demand for explainable machine learning (ML) models has been growing rapidly in recent years. Amongst the methods proposed to associate ML model predictions with human-understandable rationale, counterfactual explanations are one of the most popular. They consist of post-hoc rules derived from counterfactual examples (CFs), i.e., modified versions of input samples that result in alternative output responses from the predictive model to be explained. However, existing CF generation strategies either exploit the internals of specific models (e.g., random forests or neural networks), or depend on each sample's neighborhood, which makes them hard to be generalized for more complex models and inefficient for larger datasets. In this work, we aim to overcome these limitations and introduce a model-agnostic algorithm to generate optimal counterfactual explanations. Specifically, we formulate the problem of crafting CFs as a sequential decision-making task and then find the optimal CFs via deep reinforcement learning (DRL) with discrete-continuous hybrid action space. Differently from other techniques, our method is easily applied to any black-box model, as this resembles the environment that the DRL agent interacts with. In addition, we develop an algorithm to extract explainable decision rules from the DRL agent's policy, so as to make the process of generating CFs itself transparent. Extensive experiments conducted on several datasets have shown that our method outperforms existing CF generation baselines.

Downloads

Next from AIxIA 2021

Explanations in terms of Hierarchically organised Middle Level Features

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AIxIA 2021

Explanations in terms of Hierarchically organised Middle Level Features

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES

Downloads