A5054683542- Complex Network Analysis Techniques
- Financial Distress and Bankruptcy Prediction
- Stock Market Forecasting Methods
- Human Mobility and Location-Based Analysis
- Healthcare Operations and Scheduling Optimization
- Cardiac, Anesthesia and Surgical Outcomes
- Spam and Phishing Detection
- Misinformation and Its Impacts
- COVID-19 epidemiological studies
- Hemodynamic Monitoring and Therapy
- Knowledge Management and Sharing
- Surgical Simulation and Training
- IoT and Edge/Fog Computing
- Simulation Techniques and Applications
- Technology Adoption and User Behaviour
- Corporate Identity and Reputation
- Social Media and Politics
- Imbalanced Data Classification Techniques
- Data-Driven Disease Surveillance
- Sentiment Analysis and Opinion Mining
- Credit Risk and Financial Regulations
- Network Security and Intrusion Detection
- Data Stream Mining Techniques
- Digital Marketing and Social Media
- Advanced Graph Neural Networks
University of Parma
2018-2024
Human activity recognition (HAR) is currently recognized as a key element of more general framework designed to perform continuous monitoring human behaviors in the area ambient assisted living (AAL), well-being management, medical diagnosis, elderly care, rehabilitation, entertainment, and surveillance smart home environments. In this paper, an innovative HAR system, exploiting potential wearable devices integrated with skills deep learning techniques, presented aim recognizing most common...
Predicting corporate bankruptcy is one of the fundamental tasks in credit risk assessment. In particular, since 2007/2008 financial crisis, it has become a priority for most institutions, practitioners, and academics. The recent advancements machine learning (ML) enabled development several models prediction. challenging aspect this task dealing with class imbalance due to rarity events real economy. Furthermore, fair comparison literature difficult make because datasets are not publicly...
A troll is usually defined as somebody who provokes and offends people to make them angry, wants dominate any discussion or tries manipulate people’s opinions. The problems caused by such persons have increased with the diffusion of social media. Therefore, on one hand, press bodies magazines begun address issue write articles about phenomenon its related while, other universities research centres study features characterizing trolls look for solutions their identification. This survey aims...
ActoDatA (Actor Data Analysis) is an actor-based software library for the development of distributed data mining applications. It provides a multi-agent architecture with set predefined and configurable agents performing typical tasks In particular, its can manage different users’ applications; it maintains high level execution quality by distributing applications on dynamic computational nodes. Moreover, reports about analysis results collected data, which be accessed through either web...
This article describes the current landscape in fields of social media and socio-technical systems. In particular, it analyzes different ways which are adopted organizations, workplaces, educational smart environments. One interesting aspect this integration, is use for members’ participation access to processes services their organization. Those cover many types daily routines life activities, such as health, education, transports. survey, we compare classify research works according...
Modeling and forecasting the spread of COVID-19 remains an open problem for several reasons. One these concerns difficulty to model a complex system at high resolution (fine-grained) level which can be simulated by taking into account individual features. Agent-based modeling usually needs find optimal trade-off between simulation population size. Indeed, single individuals leads simulations smaller populations or use meta-populations. In this article, we propose solution efficiently...
Public companies in the US stock market must annually report their activities and financial performances to SEC by filing so‐called 10‐K form. Recent studies have demonstrated that changes textual content of corporate annual (10‐K) can convey strong signals companies’ future returns. In this study, we combine natural language processing techniques network science introduce a novel 10‐K‐based network, named Lazy Network, leverages year‐on‐year 10‐Ks detected using neural embedding model. The...
With the recent advances in Machine Learning (ML), several models have been successfully applied to financial and accounting data predict likelihood of companies’ bankruptcy. However, time series received little attention literature with a lack studies on application Deep sequence as Recurrent Neural Networks (RNN) Attention-based general. In this research work, we investigated Long Short Term Memory (LSTM) networks exploit for bankruptcy prediction. The main contributions our work are...
When developing a software engineering project, selecting the most appropriate programming language is crucial step. Most often, feeling at ease with possible options becomes almost as relevant technical features of language. Therefore, it appears to be worth analyzing role that emotional component plays in this process. In article, we analyze trend emotions expressed by developers 2018 on Stack Overflow platform posts concerning 26 languages. To do so, propose learning model trained distant...
In the knowledge discovery field of Big Data domain analysis geographic positioning and mobility information plays a key role. At same time, in Natural Language Processing (NLP) pre-trained models such as BERT word embedding algorithms Word2Vec enabled rich encoding words that allows mapping textual data into points an arbitrary multi-dimensional space, which notion proximity reflects association among terms or topics. The main contribution this paper is to show how analytical tools,...
Management of operating rooms is a critical point in health care organizations because surgical departments represent significant cost hospital budgets. Therefore, it increasingly important that there effective planning elective, emergency, and day surgery optimization both the human physical resources available, always maintaining high level treatment. This would lead to reduction patient waiting lists better performance not only but also entire hospital.This study aims automatically...
Abstract Modern financial markets produce massive datasets that need to be analysed using new modelling techniques like those from (deep) Machine Learning and Artificial Intelligence. The common goal of these is forecast the behaviour market, which can translated into various classification tasks, such as, for instance, predicting likelihood companies’ bankruptcy or in fraud detection systems. However, it often case real-world data are unbalanced, meaning classes’ distribution not equally...
This report presents the activities and outcomes related to five use cases (UCs) identified by EFSA within Specific Contract No 1 implementing Framework contract OC/EFSA/AMU/2021/03. UC1 aims assess usability effectiveness of text summarisation tools applied context Public consultations (PCs) with objective AI-based automatic (ATS) techniques generate a summary for each attachment comments received PC. UC2 at automating keywords identification step systematic review (SR) process, explore...
The application of Machine Learning techniques over networks, such as prediction tasks nodes and edges, is becoming often crucial in the analysis Complex systems a wide range research fields. One enabling technologies that sense represented by Node Embedding, which enables us to learn features automatically network. Among different approaches proposed literature, most promising are DeepWalk Node2Vec, where embedding computed combining random walks neural language models. However,...