Secure and Privacy-aware Data Collection and Processing in Mobile Health Systems

Healthcare systems have assimilated information and communication technologies in order to improve the quality of healthcare and patient's experience at reduced costs. The increasing digitalization of people's health information raises however new threats regarding information security and privacy. Accidental or deliberate data breaches of health data may lead to societal pressures, embarrassment and discrimination. Information security and privacy are paramount to achieve high quality healthcare services, and further, to not harm individuals when providing care. With that in mind, we give special attention to the category of Mobile Health (mHealth) systems. That is, the use of mobile devices (e.g., mobile phones, sensors, PDAs) to support medical and public health. Such systems, have been particularly successful in developing countries, taking advantage of the flourishing mobile market and the need to expand the coverage of primary healthcare programs. Many mHealth initiatives, however, fail to address security and privacy issues. This, coupled with the lack of specific legislation for privacy and data protection in these countries, increases the risk of harm to individuals. The overall objective of this thesis is to enhance knowledge regarding the design of security and privacy technologies for mHealth systems. In particular, we deal with mHealth Data Collection Systems (MDCSs), which consists of mobile devices for collecting and reporting health-related data, replacing paper-based approaches for health surveys and surveillance. This thesis consists of publications contributing to mHealth security and privacy in various ways: with a comprehensive literature review about mHealth in Brazil; with the design of a security framework for MDCSs (SecourHealth); with the design of a MDCS (GeoHealth); with the design of Privacy Impact Assessment template for MDCSs; and with the study of ontology-based obfuscation and anonymisation functions for health data.

Examining Levels of Automation in the Wood Processing Industry - A case study

Companies operating in the wood processing industry need to increase their productivity by implementing automation technologies in their production systems. An increasing global competition and rising raw material prizes challenge their competitiveness. Yet, too extensive automation brings risks such as a deterioration in situation awareness and operator deskilling. The concept of Levels of Automation is generally seen as means to achieve a balanced task allocation between the operators’ skills and competences and the need for automation technology relieving the humans from repetitive or hazardous work activities. The aim of this thesis was to examine to what extent existing methods for assessing Levels of Automation in production processes are applicable in the wood processing industry when focusing on an improved competitiveness of production systems. This was done by answering the following research questions (RQ): RQ1: What method is most appropriate to be applied with measuring Levels of Automation in the wood processing industry? RQ2: How can the measurement of Levels of Automation contribute to an improved competitiveness of the wood processing industry’s production processes? Literature reviews were used to identify the main characteristics of the wood processing industry affecting its automation potential and appropriate assessment methods for Levels of Automation in order to answer RQ1. When selecting the most suitable method, factors like the relevance to the target industry, application complexity or operational level the method is penetrating were important. The DYNAMO++ method, which covers both a rather quantitative technical-physical and a more qualitative social-cognitive dimension, was seen as most appropriate when taking into account these factors. To answer RQ 2, a case study was undertaken at a major Swedish manufacturer of interior wood products to point out paths how the measurement of Levels of Automation contributes to an improved competitiveness of the wood processing industry. The focus was on the task level on shop floor and concrete improvement suggestions were elaborated after applying the measurement method for Levels of Automation. Main aspects considered for generalization were enhancements regarding ergonomics in process design and cognitive support tools for shop-floor personnel through task standardization. Furthermore, difficulties regarding the automation of grading and sorting processes due to the heterogeneous material properties of wood argue for a suitable arrangement of human intervention options in terms of work task allocation.  The application of a modified version of DYNAMO++ reveals its pros and cons during a case study which covers a high operator involvement in the improvement process and the distinct predisposition of DYNAMO++ to be applied in an assembly system.

How to Efficiently Control the Title and Registration Mail-in Workload Backlog Processing

The Titles and Registration Mail-In Unit of the Department of Motor Vehicles only processes title and registration work that has been mailed in. During the time periods of May through June and November through January, the mail workload backlog becomes overbearing. This research will attempt to find out the cause of this increase backlog during those periods, determine what procedures or steps are currently in place and unnecessarily creating needless work that has a direct correlation with the backlog and deploy a recommendation that will totally eliminate peak time backlog work loads.

Waste Disposal Requirements for Commercial Deer-Processing Facilities

Waste from a deer-processing facility is required by law to be disposed of in any permitted municipal solid waste (Class 3) landfill or a permitted construction and demolition debris (Class 2) landfill that has specific approval to accept animal carcasses. Disposal of deer-processing waste on the property of the processing facility or at any other unpermitted site IS NOT ALLOWED. Waste from commercial deer processing facilities must be disposed of per the requirements of the S.C. Solid Waste Policy and Management Act of 1991.

Verbalizzazione di eventi biomedici espressi nella letteratura scientifica: generazione controllata di linguaggio naturale da grafi di conoscenza mediante transformer text-to-text

Il periodo in cui viviamo rappresenta la cuspide di una forte e rapida evoluzione nella comprensione del linguaggio naturale, raggiuntasi prevalentemente grazie allo sviluppo di modelli neurali. Nell'ambito dell'information extraction, tali progressi hanno recentemente consentito di riconoscere efficacemente relazioni semantiche complesse tra entità menzionate nel testo, quali proteine, sintomi e farmaci. Tale task -- reso possibile dalla modellazione ad eventi -- è fondamentale in biomedicina, dove la crescita esponenziale del numero di pubblicazioni scientifiche accresce ulteriormente il bisogno di sistemi per l'estrazione automatica delle interazioni racchiuse nei documenti testuali. La combinazione di AI simbolica e sub-simbolica può consentire l'introduzione di conoscenza strutturata nota all'interno di language model, rendendo quest'ultimi più robusti, fattuali e interpretabili. In tale contesto, la verbalizzazione di grafi è uno dei task su cui si riversano maggiori aspettative. Nonostante l'importanza di tali contributi (dallo sviluppo di chatbot alla formulazione di nuove ipotesi di ricerca), ad oggi, risultano assenti contributi capaci di verbalizzare gli eventi biomedici espressi in letteratura, apprendendo il legame tra le interazioni espresse in forma a grafo e la loro controparte testuale. La tesi propone il primo dataset altamente comprensivo su coppie evento-testo, includendo diverse sotto-aree biomediche, quali malattie infettive, ricerca oncologica e biologia molecolare. Il dataset introdotto viene usato come base per l'addestramento di modelli generativi allo stato dell'arte sul task di verbalizzazione, adottando un approccio text-to-text e illustrando una tecnica formale per la codifica di grafi evento mediante testo aumentato. Infine, si dimostra la validità degli eventi per il miglioramento delle capacità di comprensione dei modelli neurali su altri task NLP, focalizzandosi su single-document summarization e multi-task learning.

Descriptive Text Mining in ambito medico: applicazione della metodologia POIROT

Poiché la nostra conoscenza collettiva continua ad essere digitalizzata e memorizzata, diventa più difficile trovare e scoprire ciò che stiamo cercando. Abbiamo bisogno di nuovi strumenti computazionali per aiutare a organizzare, rintracciare e comprendere queste vaste quantità di informazioni. I modelli di linguaggio sono potenti strumenti che possono essere impiegati per estrarre conoscenza statisticamente significativa ed interpretabile tramite apprendimento non supervisionato, testuali o nel codice sorgente. L’obiettivo di questa tesi è impiegare una metodologia di descriptive text mining, denominata POIROT, per analizzare i rapporti medici del dataset Adverse Drug Reaction (ADE). Si vogliono stabilire delle correlazioni significative che permettano di comprendere le ragioni per cui un determinato rapporto medico fornisca o meno informazioni relative a effetti collaterali dovuti all’assunzione di determinati farmaci.

Data-driven Signal Processing per Jamming Detection in Sistemi Satellitari

In questo elaborato vengono analizzate differenti tecniche per la detection di jammer attivi e costanti in una comunicazione satellitare in uplink. Osservando un numero limitato di campioni ricevuti si vuole identificare la presenza di un jammer. A tal fine sono stati implementati i seguenti classificatori binari: support vector machine (SVM), multilayer perceptron (MLP), spectrum guarding e autoencoder. Questi algoritmi di apprendimento automatico dipendono dalle features che ricevono in ingresso, per questo motivo è stata posta particolare attenzione alla loro scelta. A tal fine, sono state confrontate le accuratezze ottenute dai detector addestrati utilizzando differenti tipologie di informazione come: i segnali grezzi nel tempo, le statistical features, le trasformate wavelet e lo spettro ciclico. I pattern prodotti dall’estrazione di queste features dai segnali satellitari possono avere dimensioni elevate, quindi, prima della detection, vengono utilizzati i seguenti algoritmi per la riduzione della dimensionalità: principal component analysis (PCA) e linear discriminant analysis (LDA). Lo scopo di tale processo non è quello di eliminare le features meno rilevanti, ma combinarle in modo da preservare al massimo l’informazione, evitando problemi di overfitting e underfitting. Le simulazioni numeriche effettuate hanno evidenziato come lo spettro ciclico sia in grado di fornire le features migliori per la detection producendo però pattern di dimensioni elevate, per questo motivo è stato necessario l’utilizzo di algoritmi di riduzione della dimensionalità. In particolare, l'algoritmo PCA è stato in grado di estrarre delle informazioni migliori rispetto a LDA, le cui accuratezze risentivano troppo del tipo di jammer utilizzato nella fase di addestramento. Infine, l’algoritmo che ha fornito le prestazioni migliori è stato il Multilayer Perceptron che ha richiesto tempi di addestramento contenuti e dei valori di accuratezza elevati.

SynBA: A contextualized Synonim-Based adversarial Attack for text classification

With the advent of high-performance computing devices, deep neural networks have gained a lot of popularity in solving many Natural Language Processing tasks. However, they are also vulnerable to adversarial attacks, which are able to modify the input text in order to mislead the target model. Adversarial attacks are a serious threat to the security of deep neural networks, and they can be used to craft adversarial examples that steer the model towards a wrong decision. In this dissertation, we propose SynBA, a novel contextualized synonym-based adversarial attack for text classification. SynBA is based on the idea of replacing words in the input text with their synonyms, which are selected according to the context of the sentence. We show that SynBA successfully generates adversarial examples that are able to fool the target model with a high success rate. We demonstrate three advantages of this proposed approach: (1) effective - it outperforms state-of-the-art attacks by semantic similarity and perturbation rate, (2) utility-preserving - it preserves semantic content, grammaticality, and correct types classified by humans, and (3) efficient - it performs attacks faster than other methods.

Hierarchical Multi-Label Text Classification in a Low-Resource Setting

In this thesis we address a multi-label hierarchical text classification problem in a low-resource setting and explore different approaches to identify the best one for our case. The goal is to train a model that classifies English school exercises according to a hierarchical taxonomy with few labeled data. The experiments made in this work employ different machine learning models and text representation techniques: CatBoost with tf-idf features, classifiers based on pre-trained models (mBERT, LASER), and SetFit, a framework for few-shot text classification. SetFit proved to be the most promising approach, achieving better performance when during training only a few labeled examples per class are available. However, this thesis does not consider all the hierarchical taxonomy, but only the first two levels: to address classification with the classes at the third level further experiments should be carried out, exploring methods for zero-shot text classification, data augmentation, and strategies to exploit the hierarchical structure of the taxonomy during training.

Fat Substitutes In Processing Of Sausages Using Piramutaba Waste.

The aim of this study was to evaluate fat substitute in processing of sausages prepared with surimi of waste from piramutaba filleting. The formulation ingredients were mixed with the fat substitutes added according to a fractional planning 2(4-1), where the independent variables, manioc starch (Ms), hydrogenated soy fat (F), texturized soybean protein (Tsp) and carrageenan (Cg) were evaluated on the responses of pH, texture (Tx), raw batter stability (RBS) and water holding capacity (WHC) of the sausage. Fat substitutes were evaluated in 11 formulations and the results showed that the greatest effects on the responses were found to Ms, F and Cg, being eliminated from the formulation Tsp. To find the best formulation for processing piramutaba sausage was made a complete factorial planning of 2(3) to evaluate the concentrations of fat substitutes in an enlarged range. The optimum condition found for fat substitutes in the sausages formulation were carrageenan (0.51%), manioc starch (1.45%) and fat (1.2%).

Central Auditory Processing Outcome After Stroke In Children.

To investigate central auditory processing in children with unilateral stroke and to verify whether the hemisphere affected by the lesion influenced auditory competence. 23 children (13 male) between 7 and 16 years old were evaluated through speech-in-noise tests (auditory closure); dichotic digit test and staggered spondaic word test (selective attention); pitch pattern and duration pattern sequence tests (temporal processing) and their results were compared with control children. Auditory competence was established according to the performance in auditory analysis ability. Was verified similar performance between groups in auditory closure ability and pronounced deficits in selective attention and temporal processing abilities. Most children with stroke showed an impaired auditory ability in a moderate degree. Children with stroke showed deficits in auditory processing and the degree of impairment was not related to the hemisphere affected by the lesion.

Temporal Auditory Processing And Phonological Awareness In Children With Benign Epilepsy With Centrotemporal Spikes.

The aim of this research was to analyze temporal auditory processing and phonological awareness in school-age children with benign childhood epilepsy with centrotemporal spikes (BECTS). Patient group (GI) consisted of 13 children diagnosed with BECTS. Control group (GII) consisted of 17 healthy children. After neurological and peripheral audiological assessment, children underwent a behavioral auditory evaluation and phonological awareness assessment. The procedures applied were: Gaps-in-Noise test (GIN), Duration Pattern test, and Phonological Awareness test (PCF). Results were compared between the groups and a correlation analysis was performed between temporal tasks and phonological awareness performance. GII performed significantly better than the children with BECTS (GI) in both GIN and Duration Pattern test (P < 0.001). GI performed significantly worse in all of the 4 categories of phonological awareness assessed: syllabic (P = 0.001), phonemic (P = 0.006), rhyme (P = 0.015) and alliteration (P = 0.010). Statistical analysis showed a significant positive correlation between the phonological awareness assessment and Duration Pattern test (P < 0.001). From the analysis of the results, it was concluded that children with BECTS may have difficulties in temporal resolution, temporal ordering, and phonological awareness skills. A correlation was observed between auditory temporal processing and phonological awareness in the suited sample.

Biofilms Of Enterococcus Faecalis And Enterococcus Faecium Isolated From The Processing Of Ricotta And The Control Of These Pathogens Through Cleaning And Sanitization Procedures.

The biofilm formation of Enterococcus faecalis and Enterococcus faecium isolated from the processing of ricotta on stainless steel coupons was evaluated, and the effect of cleaning and sanitization procedures in the control of these biofilms was determined. The formation of biofilms was observed while varying the incubation temperature (7, 25 and 39°C) and time (0, 1, 2, 4, 6 and 8days). At 7°C, the counts of E. faecalis and E. faecium were below 2log10CFU/cm(2). For the temperatures of 25 and 39°C, after 1day, the counts of E. faecalis and E. faecium were 5.75 and 6.07log10CFU/cm(2), respectively, which is characteristic of biofilm formation. The tested sanitation procedures a) acid-anionic tensioactive cleaning, b) anionic tensioactive cleaning+sanitizer and c) acid-anionic tensioactive cleaning+sanitizer were effective in removing the biofilms, reducing the counts to levels below 0.4log10CFU/cm(2). The sanitizer biguanide was the least effective, and peracetic acid was the most effective. These studies revealed the ability of enterococci to form biofilms and the importance of the cleaning step and the type of sanitizer used in sanitation processes for the effective removal of biofilms.

Pre-processing for noise detection in gene expression classification data

Due to the imprecise nature of biological experiments, biological data is often characterized by the presence of redundant and noisy data. This may be due to errors that occurred during data collection, such as contaminations in laboratorial samples. It is the case of gene expression data, where the equipments and tools currently used frequently produce noisy biological data. Machine Learning algorithms have been successfully used in gene expression data analysis. Although many Machine Learning algorithms can deal with noise, detecting and removing noisy instances from the training data set can help the induction of the target hypothesis. This paper evaluates the use of distance-based pre-processing techniques for noise detection in gene expression data classification problems. This evaluation analyzes the effectiveness of the techniques investigated in removing noisy data, measured by the accuracy obtained by different Machine Learning classifiers over the pre-processed data.