Advancing digital healthcare research and machine learning-aided knowledge extraction amidst the COVID-19 pandemic

The rapid progression of biomedical research coupled with the explosion of scientific literature has generated an exigent need for efficient and reliable systems of knowledge extraction. This dissertation contends with this challenge through a concentrated investigation of digital health, Artificial Intelligence, and specifically Machine Learning and Natural Language Processing's (NLP) potential to expedite systematic literature reviews and refine the knowledge extraction process. The surge of COVID-19 complicated the efforts of scientists, policymakers, and medical professionals in identifying pertinent articles and assessing their scientific validity. This thesis presents a substantial solution in the form of the COKE Project, an initiative that interlaces machine reading with the rigorous protocols of Evidence-Based Medicine to streamline knowledge extraction. In the framework of the COKE (“COVID-19 Knowledge Extraction framework for next-generation discovery science”) Project, this thesis aims to underscore the capacity of machine reading to create knowledge graphs from scientific texts. The project is remarkable for its innovative use of NLP techniques such as a BERT + bi-LSTM language model. This combination is employed to detect and categorize elements within medical abstracts, thereby enhancing the systematic literature review process. The COKE project's outcomes show that NLP, when used in a judiciously structured manner, can significantly reduce the time and effort required to produce medical guidelines. These findings are particularly salient during times of medical emergency, like the COVID-19 pandemic, when quick and accurate research results are critical.

Enhancing enterprise knowledge processes via cross-media extraction

In large organizations the resources needed to solve challenging problems are typically dispersed over systems within and beyond the organization, and also in different media. However, there is still the need, in knowledge environments, for extraction methods able to combine evidence for a fact from across different media. In many cases the whole is more than the sum of its parts: only when considering the different media simultaneously can enough evidence be obtained to derive facts otherwise inaccessible to the knowledge worker via traditional methods that work on each single medium separately. In this paper, we present a cross-media knowledge extraction framework specifically designed to handle large volumes of documents composed of three types of media text, images and raw data and to exploit the evidence across the media. Our goal is to improve the quality and depth of automatically extracted knowledge.

Methods of Adaptive Extraction and Analysis of Knowledge for Knowledge-base Construction and Fast Decision Making

An approach for knowledge extraction from the information arriving to the knowledge base input and also new knowledge distribution over knowledge subsets already present in the knowledge base is developed. It is also necessary to realize the knowledge transform into parameters (data) of the model for the following decision-making on the given subset. It is assumed to realize the decision-making with the fuzzy sets’ apparatus.

Semantic relations extraction from unstructured information for domain ontologies enrichment

Based in internet growth, through semantic web, together with communication speed improvement and fast development of storage device sizes, data and information volume rises considerably every day. Because of this, in the last few years there has been a growing interest in structures for formal representation with suitable characteristics, such as the possibility to organize data and information, as well as the reuse of its contents aimed for the generation of new knowledge. Controlled Vocabulary, specifically Ontologies, present themselves in the lead as one of such structures of representation with high potential. Not only allow for data representation, as well as the reuse of such data for knowledge extraction, coupled with its subsequent storage through not so complex formalisms. However, for the purpose of assuring that ontology knowledge is always up to date, they need maintenance. Ontology Learning is an area which studies the details of update and maintenance of ontologies. It is worth noting that relevant literature already presents first results on automatic maintenance of ontologies, but still in a very early stage. Human-based processes are still the current way to update and maintain an ontology, which turns this into a cumbersome task. The generation of new knowledge aimed for ontology growth can be done based in Data Mining techniques, which is an area that studies techniques for data processing, pattern discovery and knowledge extraction in IT systems. This work aims at proposing a novel semi-automatic method for knowledge extraction from unstructured data sources, using Data Mining techniques, namely through pattern discovery, focused in improving the precision of concept and its semantic relations present in an ontology. In order to verify the applicability of the proposed method, a proof of concept was developed, presenting its results, which were applied in building and construction sector.

Partager le savoir du lexicographe: extraction et modélisation ontologique des savoirs lexicographiques

Cette recherche porte sur la lexicologie, la lexicographie et l’enseignement/apprentissage du lexique. Elle s’inscrit dans le cadre du projet Modélisation ontologique des savoirs lexicographiques en vue de leur application en linguistique appliquée, surnommé Lexitation, qui est, à notre connaissance, la première tentative d’extraction des savoirs lexicographiques — i.e. connaissances déclaratives et procédurales utilisées par des lexicographes — utilisant une méthode expérimentale. Le projet repose sur le constat que les savoirs lexicographiques ont un rôle crucial à jouer en lexicologie, mais aussi en enseignement/apprentissage du lexique. Dans ce mémoire, nous décrirons les méthodes et les résultats de nos premières expérimentations, effectuées à l’aide du Think Aloud Protocol (Ericsson et Simon, 1993). Nous expliquerons l’organisation générale des expérimentations et comment les savoirs lexicographiques extraits sont modélisés pour former une ontologie. Finalement, nous discuterons des applications possibles de nos travaux en enseignement du lexique, plus particulièrement pour la formation des maîtres.

Extracting knowledge through friendsourcing: application to new financial services

Personalization has become a key factor for the success of new ICT services. However, the personal information required is not always available in a single site, but scattered in heterogeneous sources, and extracting knowledge from raw information is not an easy job. As a result, many organizations struggle to obtain knowledge on their users useful enough for their business purposes. This paper introduces a comprehensive personal data framework that opens the knowledge extraction process up to collaboration by the involvement of new actors, while enabling users to monitor and control it. The contributions have been validated in a financial services scenario where socioeconomic knowledge on some users is generated by tapping into their social network and used to assists them in raising money from their friends.

Classification of Data to Extract Knowledge from Neural Networks

A major drawback of artificial neural networks is their black-box character. Therefore, the rule extraction algorithm is becoming more and more important in explaining the extracted rules from the neural networks. In this paper, we use a method that can be used for symbolic knowledge extraction from neural networks, once they have been trained with desired function. The basis of this method is the weights of the neural network trained. This method allows knowledge extraction from neural networks with continuous inputs and output as well as rule extraction. An example of the application is showed. This example is based on the extraction of average load demand of a power plant.

Towards Unstructured Knowledge Integration in Natural Language Processing

In the last decades, Artificial Intelligence has witnessed multiple breakthroughs in deep learning. In particular, purely data-driven approaches have opened to a wide variety of successful applications due to the large availability of data. Nonetheless, the integration of prior knowledge is still required to compensate for specific issues like lack of generalization from limited data, fairness, robustness, and biases. In this thesis, we analyze the methodology of integrating knowledge into deep learning models in the field of Natural Language Processing (NLP). We start by remarking on the importance of knowledge integration. We highlight the possible shortcomings of these approaches and investigate the implications of integrating unstructured textual knowledge. We introduce Unstructured Knowledge Integration (UKI) as the process of integrating unstructured knowledge into machine learning models. We discuss UKI in the field of NLP, where knowledge is represented in a natural language format. We identify UKI as a complex process comprised of multiple sub-processes, different knowledge types, and knowledge integration properties to guarantee. We remark on the challenges of integrating unstructured textual knowledge and bridge connections with well-known research areas in NLP. We provide a unified vision of structured knowledge extraction (KE) and UKI by identifying KE as a sub-process of UKI. We investigate some challenging scenarios where structured knowledge is not a feasible prior assumption and formulate each task from the point of view of UKI. We adopt simple yet effective neural architectures and discuss the challenges of such an approach. Finally, we identify KE as a form of symbolic representation. From this perspective, we remark on the need of defining sophisticated UKI processes to verify the validity of knowledge integration. To this end, we foresee frameworks capable of combining symbolic and sub-symbolic representations for learning as a solution.

MSight: an image analysis software for liquid chromatography-mass spectrometry.

Images obtained from high-throughput mass spectrometry (MS) contain information that remains hidden when looking at a single spectrum at a time. Image processing of liquid chromatography-MS datasets can be extremely useful for quality control, experimental monitoring and knowledge extraction. The importance of imaging in differential analysis of proteomic experiments has already been established through two-dimensional gels and can now be foreseen with MS images. We present MSight, a new software designed to construct and manipulate MS images, as well as to facilitate their analysis and comparison.

BIO-INSPIRED COMPUTATIONAL TECHNIQUES APPLIED TO THE CLUSTERING AND VISUALIZATION OF SPATIO-TEMPORAL GEOSPATIAL DATA

The coverage and volume of geo-referenced datasets are extensive and incessantly¦growing. The systematic capture of geo-referenced information generates large volumes¦of spatio-temporal data to be analyzed. Clustering and visualization play a key¦role in the exploratory data analysis and the extraction of knowledge embedded in¦these data. However, new challenges in visualization and clustering are posed when¦dealing with the special characteristics of this data. For instance, its complex structures,¦large quantity of samples, variables involved in a temporal context, high dimensionality¦and large variability in cluster shapes.¦The central aim of my thesis is to propose new algorithms and methodologies for¦clustering and visualization, in order to assist the knowledge extraction from spatiotemporal¦geo-referenced data, thus improving making decision processes.¦I present two original algorithms, one for clustering: the Fuzzy Growing Hierarchical¦Self-Organizing Networks (FGHSON), and the second for exploratory visual data analysis:¦the Tree-structured Self-organizing Maps Component Planes. In addition, I present¦methodologies that combined with FGHSON and the Tree-structured SOM Component¦Planes allow the integration of space and time seamlessly and simultaneously in¦order to extract knowledge embedded in a temporal context.¦The originality of the FGHSON lies in its capability to reflect the underlying structure¦of a dataset in a hierarchical fuzzy way. A hierarchical fuzzy representation of¦clusters is crucial when data include complex structures with large variability of cluster¦shapes, variances, densities and number of clusters. The most important characteristics¦of the FGHSON include: (1) It does not require an a-priori setup of the number¦of clusters. (2) The algorithm executes several self-organizing processes in parallel.¦Hence, when dealing with large datasets the processes can be distributed reducing the¦computational cost. (3) Only three parameters are necessary to set up the algorithm.¦In the case of the Tree-structured SOM Component Planes, the novelty of this algorithm¦lies in its ability to create a structure that allows the visual exploratory data analysis¦of large high-dimensional datasets. This algorithm creates a hierarchical structure¦of Self-Organizing Map Component Planes, arranging similar variables' projections in¦the same branches of the tree. Hence, similarities on variables' behavior can be easily¦detected (e.g. local correlations, maximal and minimal values and outliers).¦Both FGHSON and the Tree-structured SOM Component Planes were applied in¦several agroecological problems proving to be very efficient in the exploratory analysis¦and clustering of spatio-temporal datasets.¦In this thesis I also tested three soft competitive learning algorithms. Two of them¦well-known non supervised soft competitive algorithms, namely the Self-Organizing¦Maps (SOMs) and the Growing Hierarchical Self-Organizing Maps (GHSOMs); and the¦third was our original contribution, the FGHSON. Although the algorithms presented¦here have been used in several areas, to my knowledge there is not any work applying¦and comparing the performance of those techniques when dealing with spatiotemporal¦geospatial data, as it is presented in this thesis.¦I propose original methodologies to explore spatio-temporal geo-referenced datasets¦through time. Our approach uses time windows to capture temporal similarities and¦variations by using the FGHSON clustering algorithm. The developed methodologies¦are used in two case studies. In the first, the objective was to find similar agroecozones¦through time and in the second one it was to find similar environmental patterns¦shifted in time.¦Several results presented in this thesis have led to new contributions to agroecological¦knowledge, for instance, in sugar cane, and blackberry production.¦Finally, in the framework of this thesis we developed several software tools: (1)¦a Matlab toolbox that implements the FGHSON algorithm, and (2) a program called¦BIS (Bio-inspired Identification of Similar agroecozones) an interactive graphical user¦interface tool which integrates the FGHSON algorithm with Google Earth in order to¦show zones with similar agroecological characteristics.

Sensor communication in Smart cities and regions: An efficient IoT-based remote health monitoring system

Recent advances in Information and Communication Technology (ICT), especially those related to the Internet of Things (IoT), are facilitating smart regions. Among many services that a smart region can offer, remote health monitoring is a typical application of IoT paradigm. It offers the ability to continuously monitor and collect health-related data from a person, and transmit the data to a remote entity (for example, a healthcare service provider) for further processing and knowledge extraction. An IoT-based remote health monitoring system can be beneficial in rural areas belonging to the smart region where people have limited access to regular healthcare services. The same system can be beneficial in urban areas where hospitals can be overcrowded and where it may take substantial time to avail healthcare. However, this system may generate a large amount of data. In order to realize an efficient IoT-based remote health monitoring system, it is imperative to study the network communication needs of such a system; in particular the bandwidth requirements and the volume of generated data. The thesis studies a commercial product for remote health monitoring in Skellefteå, Sweden. Based on the results obtained via the commercial product, the thesis identified the key network-related requirements of a typical remote health monitoring system in terms of real-time event update, bandwidth requirements and data generation. Furthermore, the thesis has proposed an architecture called IReHMo - an IoT-based remote health monitoring architecture. This architecture allows users to incorporate several types of IoT devices to extend the sensing capabilities of the system. Using IReHMo, several IoT communication protocols such as HTTP, MQTT and CoAP has been evaluated and compared against each other. Results showed that CoAP is the most efficient protocol to transmit small size healthcare data to the remote servers. The combination of IReHMo and CoAP significantly reduced the required bandwidth as well as the volume of generated data (up to 56 percent) compared to the commercial product. Finally, the thesis conducted a scalability analysis, to determine the feasibility of deploying the combination of IReHMo and CoAP in large numbers in regions in north Sweden.

Modelling ecological values in heterogeneous and dynamic landscapes with geospatial data

Our surrounding landscape is in a constantly dynamic state, but recently the rate of changes and their effects on the environment have considerably increased. In terms of the impact on nature, this development has not been entirely positive, but has rather caused a decline in valuable species, habitats, and general biodiversity. Regardless of recognizing the problem and its high importance, plans and actions of how to stop the detrimental development are largely lacking. This partly originates from a lack of genuine will, but is also due to difficulties in detecting many valuable landscape components and their consequent neglect. To support knowledge extraction, various digital environmental data sources may be of substantial help, but only if all the relevant background factors are known and the data is processed in a suitable way. This dissertation concentrates on detecting ecologically valuable landscape components by using geospatial data sources, and applies this knowledge to support spatial planning and management activities. In other words, the focus is on observing regionally valuable species, habitats, and biotopes with GIS and remote sensing data, using suitable methods for their analysis. Primary emphasis is given to the hemiboreal vegetation zone and the drastic decline in its semi-natural grasslands, which were created by a long trajectory of traditional grazing and management activities. However, the applied perspective is largely methodological, and allows for the application of the obtained results in various contexts. Models based on statistical dependencies and correlations of multiple variables, which are able to extract desired properties from a large mass of initial data, are emphasized in the dissertation. In addition, the papers included combine several data sets from different sources and dates together, with the aim of detecting a wider range of environmental characteristics, as well as pointing out their temporal dynamics. The results of the dissertation emphasise the multidimensionality and dynamics of landscapes, which need to be understood in order to be able to recognise their ecologically valuable components. This not only requires knowledge about the emergence of these components and an understanding of the used data, but also the need to focus the observations on minute details that are able to indicate the existence of fragmented and partly overlapping landscape targets. In addition, this pinpoints the fact that most of the existing classifications are too generalised as such to provide all the required details, but they can be utilized at various steps along a longer processing chain. The dissertation also emphases the importance of landscape history as an important factor, which both creates and preserves ecological values, and which sets an essential standpoint for understanding the present landscape characteristics. The obtained results are significant both in terms of preserving semi-natural grasslands, as well as general methodological development, giving support to science-based framework in order to evaluate ecological values and guide spatial planning.

Étude comparative du vocabulaire de description de la danse dans les archives et du vocabulaire de représentation de la danse dans la littérature

Notre recherche s’insère dans la mouvance des humanités numériques; nous y faisons dialoguer les arts et les sciences de l’information. Depuis quelques décennies, la danse est un sujet d’études et de recherche à part entière. Il devient donc nécessaire de mieux décrire la danse dans les archives, sachant que la description en amont influe grandement sur l’accès en aval. Les méthodes d’extraction automatique de connaissances nous semblent offrir de nouvelles possibilités. L’objectif de notre recherche est de contribuer au développement d’outils de gestion de l’information dans les archives de la danse en comparant un vocabulaire de description de la danse dans les archives et un vocabulaire de représentation de la danse dans la littérature, recueilli grâce à des méthodes d’extraction automatique de connaissances, pour en distinguer une possible complémentarité, particulièrement en ce qui a trait au vocabulaire de l’expérience esthétique. D’abord, nous analysons un vocabulaire de description de la danse dans les archives. Nous décrivons certains outils de description des archives de la danse et nous analysons le thésaurus de descripteurs Collier. Nous constatons que le vocabulaire de description de la danse dans les archives ne semble pas prendre en compte l’expérience esthétique. Ensuite, nous analysons un vocabulaire de représentation de la danse dans la littérature. Un vocabulaire structuré de l’expérience esthétique de la danse moderne est ainsi extrait d’un corpus de textes de l’écrivain français Stéphane Mallarmé et analysé. Puis nous comparons les deux vocabulaires afin d'en distinguer la complémentarité quant à la description de l’expérience esthétique. Nous formulons une première suggestion d’amélioration de certains thésaurus employés dans les archives de la danse : un thésaurus au vocabulaire essentiellement factuel, comme le thésaurus de descripteurs Collier, peut être enrichi de termes à propos de l’expérience esthétique. Le vocabulaire de représentation de la danse dans la littérature est jusqu’à un certain point complémentaire au vocabulaire de description de l’expérience esthétique de la danse dans les archives. Nous menons ainsi une première expérimentation qui justifie en partie la pertinence de certaines méthodes d’extraction de connaissances dans le développement et la maintenance de ressources documentaires pour le domaine des arts d’interprétation tels que la danse.

Conceptual design of wastewater treatment plants using multiple objectives

La implementació de la Directiva Europea 91/271/CEE referent a tractament d'aigües residuals urbanes va promoure la construcció de noves instal·lacions al mateix temps que la introducció de noves tecnologies per tractar nutrients en àrees designades com a sensibles. Tant el disseny d'aquestes noves infraestructures com el redisseny de les ja existents es va portar a terme a partir d'aproximacions basades fonamentalment en objectius econòmics degut a la necessitat d'acabar les obres en un període de temps relativament curt. Aquests estudis estaven basats en coneixement heurístic o correlacions numèriques provinents de models determinístics simplificats. Així doncs, moltes de les estacions depuradores d'aigües residuals (EDARs) resultants van estar caracteritzades per una manca de robustesa i flexibilitat, poca controlabilitat, amb freqüents problemes microbiològics de separació de sòlids en el decantador secundari, elevats costos d'operació i eliminació parcial de nutrients allunyant-les de l'òptim de funcionament. Molts d'aquestes problemes van sorgir degut a un disseny inadequat, de manera que la comunitat científica es va adonar de la importància de les etapes inicials de disseny conceptual. Precisament per aquesta raó, els mètodes tradicionals de disseny han d'evolucionar cap a sistemes d'avaluació mes complexos, que tinguin en compte múltiples objectius, assegurant així un millor funcionament de la planta. Tot i la importància del disseny conceptual tenint en compte múltiples objectius, encara hi ha un buit important en la literatura científica tractant aquest camp d'investigació. L'objectiu que persegueix aquesta tesi és el de desenvolupar un mètode de disseny conceptual d'EDARs considerant múltiples objectius, de manera que serveixi d'eina de suport a la presa de decisions al seleccionar la millor alternativa entre diferents opcions de disseny. Aquest treball de recerca contribueix amb un mètode de disseny modular i evolutiu que combina diferent tècniques com: el procés de decisió jeràrquic, anàlisi multicriteri, optimació preliminar multiobjectiu basada en anàlisi de sensibilitat, tècniques d'extracció de coneixement i mineria de dades, anàlisi multivariant i anàlisi d'incertesa a partir de simulacions de Monte Carlo. Això s'ha aconseguit subdividint el mètode de disseny desenvolupat en aquesta tesis en quatre blocs principals: (1) generació jeràrquica i anàlisi multicriteri d'alternatives, (2) anàlisi de decisions crítiques, (3) anàlisi multivariant i (4) anàlisi d'incertesa. El primer dels blocs combina un procés de decisió jeràrquic amb anàlisi multicriteri. El procés de decisió jeràrquic subdivideix el disseny conceptual en una sèrie de qüestions mes fàcilment analitzables i avaluables mentre que l'anàlisi multicriteri permet la consideració de diferent objectius al mateix temps. D'aquesta manera es redueix el nombre d'alternatives a avaluar i fa que el futur disseny i operació de la planta estigui influenciat per aspectes ambientals, econòmics, tècnics i legals. Finalment aquest bloc inclou una anàlisi de sensibilitat dels pesos que proporciona informació de com varien les diferents alternatives al mateix temps que canvia la importància relativa del objectius de disseny. El segon bloc engloba tècniques d'anàlisi de sensibilitat, optimització preliminar multiobjectiu i extracció de coneixement per donar suport al disseny conceptual d'EDAR, seleccionant la millor alternativa un cop s'han identificat decisions crítiques. Les decisions crítiques són aquelles en les que s'ha de seleccionar entre alternatives que compleixen de forma similar els objectius de disseny però amb diferents implicacions pel que respecte a la futura estructura i operació de la planta. Aquest tipus d'anàlisi proporciona una visió més àmplia de l'espai de disseny i permet identificar direccions desitjables (o indesitjables) cap on el procés de disseny pot derivar. El tercer bloc de la tesi proporciona l'anàlisi multivariant de les matrius multicriteri obtingudes durant l'avaluació de les alternatives de disseny. Específicament, les tècniques utilitzades en aquest treball de recerca engloben: 1) anàlisi de conglomerats, 2) anàlisi de components principals/anàlisi factorial i 3) anàlisi discriminant. Com a resultat és possible un millor accés a les dades per realitzar la selecció de les alternatives, proporcionant més informació per a una avaluació mes efectiva, i finalment incrementant el coneixement del procés d'avaluació de les alternatives de disseny generades. En el quart i últim bloc desenvolupat en aquesta tesi, les diferents alternatives de disseny són avaluades amb incertesa. L'objectiu d'aquest bloc és el d'estudiar el canvi en la presa de decisions quan una alternativa és avaluada incloent o no incertesa en els paràmetres dels models que descriuen el seu comportament. La incertesa en el paràmetres del model s'introdueix a partir de funcions de probabilitat. Desprès es porten a terme simulacions Monte Carlo, on d'aquestes distribucions se n'extrauen números aleatoris que es subsisteixen pels paràmetres del model i permeten estudiar com la incertesa es propaga a través del model. Així és possible analitzar la variació en l'acompliment global dels objectius de disseny per a cada una de les alternatives, quines són les contribucions en aquesta variació que hi tenen els aspectes ambientals, legals, econòmics i tècnics, i finalment el canvi en la selecció d'alternatives quan hi ha una variació de la importància relativa dels objectius de disseny. En comparació amb les aproximacions tradicionals de disseny, el mètode desenvolupat en aquesta tesi adreça problemes de disseny/redisseny tenint en compte múltiples objectius i múltiples criteris. Al mateix temps, el procés de presa de decisions mostra de forma objectiva, transparent i sistemàtica el perquè una alternativa és seleccionada en front de les altres, proporcionant l'opció que més bé acompleix els objectius marcats, mostrant els punts forts i febles, les principals correlacions entre objectius i alternatives, i finalment tenint en compte la possible incertesa inherent en els paràmetres del model que es fan servir durant les anàlisis. Les possibilitats del mètode desenvolupat es demostren en aquesta tesi a partir de diferents casos d'estudi: selecció del tipus d'eliminació biològica de nitrogen (cas d'estudi # 1), optimització d'una estratègia de control (cas d'estudi # 2), redisseny d'una planta per aconseguir eliminació simultània de carboni, nitrogen i fòsfor (cas d'estudi # 3) i finalment anàlisi d'estratègies control a nivell de planta (casos d'estudi # 4 i # 5).