In a complex context with many classes and instances, synthetic classification evaluation techniques (e.g. classical measures made from confusion matrix) often obscure important information. Such information usually enables experts to achieve well-informed decisions, particularly under strong temporal constraints. Visual analysis thus appears to be a promising additional support. We thus propose an approach to visualize the results of a multi-class classifier at instance level in the form of a bar chart and a heatmap by class. The novelty of this approach lies in presenting visually to experts the entirety of the classifier's predictions, allowing them to synthesize these outcomes by themselves. By doing so, they can dynamically adjust their confidence levels and make informed decisions. We identify several user objectives and illustrate the interest and validity of our approach on AIS (Automatic Identification System) radio frequency fingerprinting.

Introduction:Peculiar use of language and even language deficits are one of the well-known signs of schizophrenia. Different language features analyzed using natural language processing and machine learning have been reported to differentiate patients at ultra-high risk for psychosis. However, it has not always been explained how, and to what extent, those linguistic markers allow the distinction of patients. This study aims to find relevant linguistic markers for classifying patients at ultra-high risk and explain how the detected markers contribute to the classification. Methods:The first consultations with a psychiatrist of 68 patients (15 not-at-risk patients, 45 at-risk patients, and 8 patients with first episode psychosis) were recorded, transcribed verbatim, and annotated for analyses using natural language processing. A gradient-boosted decision tree algorithm was tested to evaluate its potential to correctly classify three categories of patients and find relevant linguistic markers at the level of lexical richness, semantic coherence, speech disfluency, and syntactic complexity. The Synthetic Minority Oversampling Technique was used to handle imbalanced data, and the SHapley Additive exPlanations (SHAP) values were computed to measure feature importance and each feature's contributions to the classification. Results:The model yielded good performance, that is, 0.82 accuracy, 0.82 F2-score, 0.85 precision, 0.82 recall, and 0.86 ROC-AUC score, with four linguistic variables that concern weak coherence, the use of "I," and filled pauses. Discussion:The findings in this study suggest that weak coherence play a key role in classification. No significant differences in the use of "I" and filled pauses were found between groups using a statistical test, but an explainability model showed its different contributions. The contribution of each linguistic feature to the classification by patient group provided deeper insight into linguistic manifestations of each patient group and their subtle differences, which could help better analyze and understand patients' language behaviors.

Bipolar disorder is one of the ten most incapacitating disorders, according to the World Health Organization, and its early diagnosis is crucial to avoid disease-related complications. Language offers a rich data source for analyzing mental disorders, and the analysis of acoustic parameters, syntax, semantic content, and language structure, in general, is a constantly evolving field. In this study, we conducted a systematic review of the literature to identify markers in the oral language of mania and hypomania in bipolar disorder, using recordings of clinical interviews. We conducted our study using the EMBASE, Pubmed, IEEE, ACL Anthology, Scopus and arXiv databases. Results show several linguistic and acoustic variations in manic and hypomanic episodes, including verbal overproduction, decreased pauses, unstable speech, and communication disorders. The studies also reveal significant spectral and energetic variations. The data point to the existence of changes in the speech characteristics of patients with bipolar disorder in the manic or hypomanic phase. Promising results have been obtained. However, several challenges remain, and further studies are still needed to provide significant advance in diagnostic for bipolar disorder.

Motivated by the distinction between semantics and pragmatics as sub-disciplines of linguistics, shortly after Tim Berners-Lee introduced the Semantic Web in 2001, there have been works on its extension to the "pragmatic level". Twenty years later, the Semantic Web is more popular than ever, while little has been achieved in extending it into a Pragmatic Web. Social representations introduced by Serge Moscovici in the 1960s seem totally ignored by the information technology community even though they are strongly related to research on opinion mining and representation in social media. We, thus, recall the major results of academic research on the Pragmatic Web, followed by our proposal for an Implicit Pragmatic Web inspired by various sub-domains of the discipline of pragmatics. We further recall the basics of the social representations theory and discuss their potential implementations in a Web of Social Representations and thus their potential contribution towards at least a part of the future internet.

Complex systems represented by multivariate time series are ubiquitous in many applications, especially in industry.Understanding a complex system, its states and their evolution over time is a challenging task.This is due to the permanent change of contextual events internal and external to the system.We are interested in representing the evolution of a complex system in an intelligible and explainable way based on knowledge extraction.We propose XR-CSB (eXplainable Representation of Complex System Behavior) based on three steps: (i) a time series vertical clustering to detect system states, (ii) an explainable visual representation using unfolded finite-state automata and (iii) an explainable pre-modeling based on an enrichment via exploratory metrics.Four representations adapted to the expertise level of domain experts for acceptability issues are proposed.Experiments show that XR-CSB is scalable.Qualitative evaluation by experts of different expertise levels shows that XR-CSB meets their expectations in terms of explainability, intelligibility and acceptability.

Dealing with redundancy is one of the main challenges in frequency based data mining and itemset mining in particular. To tackle this issue in the most objective possible way, we introduce the theoretical bases of a new probabilistic concept: Mutual constrained independence (MCI). Thanks to this notion, we describe a MCI model for the frequencies of all itemsets which is the least binding in terms of model hypotheses defined by the knowledge of the frequencies of some of the itemsets. We provide a method for computing MCI models based on algebraic geometry. We establish the link between MCI models and a class of MaxEnt models which has already known to be used in pattern mining. As such, our research presents further insight on the nature of such models and an entirely novel approach for computing them. (C) 2021 Elsevier Inc. All rights reserved.

Graph embedding aims to learn a representation of graphs' nodes in a latent low-dimensional space. The purpose is to encode the graph's structural information. While the majority of real-world networks are dynamic, literature generally focuses on static networks and overlooks evolution patterns. In a previous article entitled TemporalNode2vec: Temporal Node Embedding in Temporal Networks, we introduced a dynamic graph embedding method that learns continuous time-aware vertex representations. In this paper, we adapt TemporalNode2vec to tackle especially the node classification-related tasks. Overall, we prove that task-specific embedding improves data efficiency significantly comparing to task-agnostic embedding.

Graph autoencoders (GAE), also known as graph embedding methods, learn latent representations of the nodes of a graph in a low-dimensional space where the structural information is preserved. While real-world graphs are generally dynamic, only a few embedding methods handle the temporal dimension: Even though they have proven their reliability, the majority of the embedding techniques address the case of static networks and present poor performances when applied to temporal ones. In this paper, we present a generic method to temporalize static graph autoencoders, i.e. adapt different static graph embedding methods to the case of temporal networks. This is made possible by learning optimal connections between timesteps' graphs in order to form a single merged spatio-temporal network. We prove that this highly improves the inference tasks' accuracy of the temporalized methods. We also show that the learned connections are directly related to nodes characteristics and can be used beyond the scope of the embedding they are designed for.

Industrial Control Systems (ICS) or SCADA networks are increasingly targeted by cyber-attacks as their architectures shifted from proprietary hardware, software and protocols to standard and open sources ones.Furthermore, these systems which used to be isolated are now interconnected to corporate networks and to the Internet.Among the countermeasures to mitigate the threats, anomaly detection systems play an important role as they can help detect even unknown attacks.Deep learning which has gained a great attention in the last few years due to excellent results in image, video and natural language processing is being used for anomaly detection in information security, particularly in SCADA networks.The salient features of the data from SCADA networks are learnt as hierarchical representation using deep architectures, and those learnt features are used to classify the data into normal or anomalous ones.This article is a review of various architectures such as Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Stacked Autoencoder (SAE), Long Short Term Memory (LSTM), or a combination of those architectures, for anomaly detection purpose in SCADA networks.

Presentation in the context of the Twelfth International Workshop Modelling and Reasoning in Context (MRC) @IJCAI 2021, of the following article : Chraibi Kaadoud, Ikram, Lina Fahed, and Philippe Lenca. Explainable AI: a narrative review at the crossroad of Knowledge Discovery, Knowledge Representation and Representation Twelfth International Workshop Modelling and Reasoning in Context. 2021. ⟨hal-03343687⟩ Abstract: EXplainable Artificial Intelligence (XAI) has recently become a very active domain, mainly due to the extensive development of black-box models such as neural networks. Recent XAI objectives have been defined in the state-of-the-art, for which specific approaches have been proposed. Implicit links can be found between XAI and other domains, especially related to knowledge and neural networks. We here aim to highlight these implicit links. We present a narrative review of research works in two domains: (i) Knowledge domain with focus on Knowledge Discovery and Representation, and (ii) Representation Learning. We discuss the similarity and joining points between these domains and XAI. We conclude that, in order to make black-boxes more transparent, XAI approaches should be more inspired and take advantage of past and recent works in Knowledge and Representation Learning domains. Through this paper, we offer an entry point to the domain of XAI for both multidisciplinary researchers and specialists in AI, as well for AI knowledgeable users. Link to article: TBC Link to the workshop: http://mrc.kriwi.de/hccs2021/ Link to IJCAI conference: https://ijcai-21.org/ About the authors: Ikram Chraibi Kaadoud, XAI Postdoctoral Researcher: https://www.imt-atlantique.fr/en/person/ikram-chraibi-kaadoud Lina Fahed, Associate professor in Computer Science: https://www.imt-atlantique.fr/en/person/lina-fahed Philippe Lenca, Professor at IMT Atlantique, Head of Logics in Uses, Social Science and Information Science department.: https://www.imt-atlantique.fr/en/person/philippe-lenca Research team: DECIDE Team : https://www.labsticc.fr/en/teams/m-570-decide.htm LAB-STICC laboratory : https://www.labsticc.fr/en/index/ LUSSI Department : https://www.imt-atlantique.fr/fr/l-ecole/departements-d-enseignement-recherche/lussi

SCADA systems are more and more targeted by cyber-attacks because of many vulnerabilities inhardware, software, protocols and the communication stack. Those systems nowadays use standard hardware, software, operating systems and protocols. Furthermore, SCADA systems which used to be air-gaped are now interconnected to corporate networks and to the Internet, widening the attack surface.In this thesis, we are using a deep learning approach to propose an efficient hybrid deep neural network for anomaly detection in SCADA systems. The salient features of SCADA data are automatically and unsupervisingly learnt, and then fed to a supervised classifier in order to dertermine if those data are normal or abnormal, i.e if there is a cyber-attack or not. Afterwards, as a response to the challenge caused by high training time of deep learning models, we proposed a distributed approach of our anomaly detection system in order lo lessen the training time of our model.

BACKGROUND Machine learning systems are part of the field of artificial intelligence that automatically learn models from data to make better decisions. Natural language processing (NLP), by using corpora and learning approaches, provides good performance in statistical tasks, such as text classification or sentiment mining. OBJECTIVE The primary aim of this systematic review was to summarize and characterize, in methodological and technical terms, studies that used machine learning and NLP techniques for mental health. The secondary aim was to consider the potential use of these methods in mental health clinical practice METHODS This systematic review follows the PRISMA (Preferred Reporting Items for Systematic Review and Meta-analysis) guidelines and is registered with PROSPERO (Prospective Register of Systematic Reviews; number CRD42019107376). The search was conducted using 4 medical databases (PubMed, Scopus, ScienceDirect, and PsycINFO) with the following keywords: machine learning, data mining, psychiatry, mental health, and mental disorder. The exclusion criteria were as follows: languages other than English, anonymization process, case studies, conference papers, and reviews. No limitations on publication dates were imposed. RESULTS A total of 327 articles were identified, of which 269 (82.3%) were excluded and 58 (17.7%) were included in the review. The results were organized through a qualitative perspective. Although studies had heterogeneous topics and methods, some themes emerged. Population studies could be grouped into 3 categories: patients included in medical databases, patients who came to the emergency room, and social media users. The main objectives were to extract symptoms, classify severity of illness, compare therapy effectiveness, provide psychopathological clues, and challenge the current nosography. Medical records and social media were the 2 major data sources. With regard to the methods used, preprocessing used the standard methods of NLP and unique identifier extraction dedicated to medical texts. Efficient classifiers were preferred rather than transparent functioning classifiers. Python was the most frequently used platform. CONCLUSIONS Machine learning and NLP models have been highly topical issues in medicine in recent years and may be considered a new paradigm in medical research. However, these processes tend to confirm clinical hypotheses rather than developing entirely new information, and only one major category of the population (ie, social media users) is an imprecise cohort. Moreover, some language-specific features can improve the performance of NLP methods, and their extension to other languages should be more closely investigated. However, machine learning and NLP techniques provide useful information from unexplored data (ie, patients’ daily habits that are usually inaccessible to care providers). Before considering It as an additional tool of mental health care, ethical issues remain and should be discussed in a timely manner. Machine learning and NLP methods may offer multiple perspectives in mental health research but should also be considered as tools to support clinical practice.

Community detection emerges as an important task in the discovery of network mesoscopic structures. However, the concept of a “good” community is very context-dependent, and it is relatively complicated to deduce community characteristics using available community detection techniques. In reality, the existence of a gap between structural goodness quality metrics and expected topological patterns creates a confusion in evaluating community structures. We thus introduce an empirical multivariate analysis of different structural goodness properties in order to characterize several detectable community topologies. Specifically, we show that a combination of two representative structural dimensions including community transitivity and hub dominance allows to distinguish different topologies such as star-based, clique-based, string-based and grid-based structures. Additionally, these classes of topology disclose structural proximities with those of graphs created by Erdős–Rényi, Watts–Strogatz and Barabási–Albert generative models. We illustrate popular community topologies identified by different detection methods on a large dataset composing many network categories and associate their structures with the most related graph generative model. Interestingly, this conjunctive representation sheds light on fundamental differences between mesoscopic structures in various network categories including communication, information, biological, technological, social, ecological, synthetic networks and more.