Perspectives in Earthworm Molecular Phylogeny: Recent Advances in Lumbricoidea and Standing Questions

Earthworm systematics have been limited by the small number of taxonomically informative morphological characters and high levels of homoplasy in this group. However, molecular phylogenetic techniques have yielded significant improvements in earthworm taxonomy in the last 15 years. Several different approaches based on the use of different molecular markers, sequencing techniques, and compromises between specimen/taxon coverage and phylogenetic information have recently emerged (DNA barcoding, multigene phylogenetics, mitochondrial genome analysis, transcriptome analysis, targeted enrichment methods, and reduced representation techniques), providing solutions to different evolutionary questions regarding European earthworms. Molecular phylogenetics have led to significant advances being made in Lumbricidae systematics, such as the redefinition or discovery of new genera (Galiciandrilus, Compostelandrilus, Vindoboscolex, Castellodrilus), delimitation and revision of previously existing genera (Kritodrilus, Eophila, Zophoscolex, Bimastos), and changes to the status of subspecific taxa (such as the Allolobophorachaetophora complex). These approaches have enabled the identification of problems that can be resolved by molecular phylogenetics, including the revision of Aporrectodea, Allolobophora, Helodrilus, and Dendrobaena, as well as the examination of small taxa such as Perelia, Eumenescolex, and Iberoscolex. Similar advances have been made with the family Hormogastridae, in which integrative systematics have contributed to the description of several new species, including the delimitation of (formerly) cryptic species. At the family level, integrative systematics have provided a new genus system that better reflects the diversity and biogeography of these earthworms, and phylogenetic comparative methods provide insight into earthworm macroevolution. Despite these achievements, further research should be performed on the Tyrrhenian cryptic complexes, which are of special eco-evolutionary interest. These examples highlight the potential value of applying molecular phylogenetic techniques to other earthworm families, which are very diverse and occupy different terrestrial habitats across the world. The systematic implementation of such approaches should be encouraged among the different expert groups worldwide, with emphasis on collaboration and cooperation.

Semantic Interoperability in Internet of Things

The industry's internet of things (IoT) applications have drawn significant research attention in recent decades. IoT is a technology in which intelligent objects with sensors-enabled RFID tags, actuators, and processors communicate information to cater to a meaningful purpose in the industry. This way, IoT technology aims to simplify the distributed data collection in industrial practice, sharing and processing information and knowledge across many collaborating partners using suitable enterprise information systems. This chapter describes new methods with grounded knowledge representation techniques to address the needs of formal information modeling and reasoning for web-based services. The chapter presents a framework, apparel business decentralized data integration (ABDDI), which uses knowledge representation methods and formal languages (e.g., description logics – DLs) to annotate necessary business activities. This type of web service requires increased interoperability in service management operations.

Multi-Keyword Classification: A Case Study in Finnish Social Sciences Data Archive

In this paper, we consider the task of assigning relevant labels to studies in the social science domain. Manual labelling is an expensive process and prone to human error. Various multi-label text classification machine learning approaches have been proposed to resolve this problem. We introduce a dataset obtained from the Finnish Social Science Archive and comprised of 2968 research studies’ metadata. The metadata of each study includes attributes, such as the “abstract” and the “set of labels”. We used the Bag of Words (BoW), TF-IDF term weighting and pretrained word embeddings obtained from FastText and BERT models to generate the text representations for each study’s abstract field. Our selection of multi-label classification methods includes a Naive approach, Multi-label k Nearest Neighbours (ML-kNN), Multi-Label Random Forest (ML-RF), X-BERT and Parabel. The methods were combined with the text representation techniques and their performance was evaluated on our dataset. We measured the classification accuracy of the combinations using Precision, Recall and F1 metrics. In addition, we used the Normalized Discounted Cumulative Gain to measure the label ranking performance of the selected methods combined with the text representation techniques. The results showed that the ML-RF model achieved a higher classification accuracy with the TF-IDF features and, based on the ranking score, the Parabel model outperformed the other methods.

A Comparative Study of Time Frequency Representation Techniques for Freeze of Gait Detection and Prediction

Freezing of Gait (FOG) is an impairment that affects the majority of patients in the advanced stages of Parkinson’s Disease (PD). FOG can lead to sudden falls and injuries, negatively impacting the quality of life for the patients and their families. Rhythmic Auditory Stimulation (RAS) can be used to help patients recover from FOG and resume normal gait. RAS might be ineffective due to the latency between the start of a FOG event, its detection and initialization of RAS. We propose a system capable of both FOG prediction and detection using signals from tri-axial accelerometer sensors that will be useful in initializing RAS with minimal latency. We compared the performance of several time frequency analysis techniques, including moving windows extracted from the signals, handcrafted features, Recurrence Plots (RP), Short Time Fourier Transform (STFT), Discreet Wavelet Transform (DWT) and Pseudo Wigner Ville Distribution (PWVD) with Deep Learning (DL) based Long Short Term Memory (LSTM) and Convolutional Neural Networks (CNN). We also propose three Ensemble Network Architectures that combine all the time frequency representations and DL architectures. Experimental results show that our ensemble architectures significantly improve the performance compared with existing techniques. We also present the results of applying our method trained on a publicly available dataset to data collected from patients using wearable sensors in collaboration with A.T. Still University.

A Comparative Study of Time Frequency Representation Techniques for Freeze of Gait Detection and Prediction

Freezing of Gait (FOG) is an impairment that affects the majority of patients in the advanced stages of Parkinson’s Disease (PD). FOG can lead to sudden falls and injuries, negatively impacting the quality of life for the patients and their families. Rhythmic Auditory Stimulation (RAS) can be used to help patients recover from FOG and resume normal gait. RAS might be ineffective due to the latency between the start of a FOG event, it’s detection and initialization of RAS. We propose a system capable of both FOG prediction and detection using signals from tri-axial accelerometer sensors that will be useful in initializing RAS with minimal latency. We compared the performance of several time frequency analysis techniques, including moving windows extracted from the signals, handcrafted features, Recurrence Plots (RP), Short Time Fourier Transform (STFT), Discreet Wavelet Transform (DWT) and Pseudo Wigner Ville Distribution (PWVD) with Deep Learning (DL) based Long Short Term Memory (LSTM) and Convolutional Neural Networks (CNN). We also propose three Ensemble Network Architectures that combine all the time frequency representations and DL architectures. Experimental results show that our ensemble architectures significantly improve the performance compared with existing techniques. We also present the results of applying our method trained on publicly available dataset to data collected from patients using wearable sensors in collaboration with A.T. Still University.

Sensor placement for sound field reconstruction in enclosures.

Sampling spatio-temporal acoustic fields is a challenging problem since it demands a large number of sensors. Typically, to characterize the pressure field inside an enclosure, the number of measurements required increases linearly with frequency and cubically with volume, becoming an intractable problem for rooms of moderate size even at low and mid frequencies. Sparse representation techniques, such as Compressed Sensing, rely on the sparsity of natural signals in certain representation domain to drastically reduce the number of measurements needed to sample such signals. In this study, we optimize the placement of sensors inside an enclosure in order to reduce the measurements required for a given reconstruction accuracy. The proposed methodology selects a sparse set of sensor positions from predefined grid via the QR factorization of the sensing matrix. Numerical results show an effective reduction in the required number of measurements when their positions are optimized, in contrast to standard random positioning. Unlike the majority of existing approaches, we study the placement problem for wide-band acoustic fields.

A MODELLING FRAMEWORK FOR DATA-DRIVEN DESIGN FOR SUSTAINABLE BEHAVIOUR IN HUMAN-MACHINE INTERACTIONS

As the society is already permeated by data, a data-driven approach to inform design for sustainable behaviour can help to identify misbehaviours and target sustainable behaviours to achieve, as well as to select and implement the most suitable design strategies to promote a behavioural change and monitor their effectiveness. This work addresses the open challenge of providing designers with a model for Human-Machine Interactions (HMI) that helps to identify relevant data to collect for inferring user behaviour related to environmental sustainability during product use.We propose a systematic modelling framework that combines constructs from existing representation techniques to identify the most critical variables for resources consumption, which are the determinants of potential misbehaviours related to HMI. The analysis is represented as a Behaviour-Inefficiency Model that graphically supports the analyst/designer to link user behaviours with a quantitative representation of resources consumption.The paper describes the model through an example of the use of a kettle and an additional application of the same approach to a washing machine, in order to point out its versatility for modelling more complex interactions.

User Representation Learning for Social Networks: An Empirical Study

Gathering useful insights from social media data has gained great interest over the recent years. User representation can be a key task in mining publicly available user-generated rich content offered by the social media platforms. The way to automatically create meaningful observations about users of a social network is to obtain real-valued vectors for the users with user embedding representation learning models. In this study, we presented one of the most comprehensive studies in the literature in terms of learning high-quality social media user representations by leveraging state-of-the-art text representation approaches. We proposed a novel doc2vec-based representation method, which can encode both textual and non-textual information of a social media user into a low dimensional vector. In addition, various experiments were performed for investigating the performance of text representation techniques and concepts including word2vec, doc2vec, Glove, NumberBatch, FastText, BERT, ELMO, and TF-IDF. We also shared a new social media dataset comprising data from 500 manually selected Twitter users of five predefined groups. The dataset contains different activity data such as comment, retweet, like, location, as well as the actual tweets composed by the users.

Screening of sleep movement disorders via EMG signal analysis

It is estimated that 50 to 70 million Americans suffer from a chronic sleep disorder, which hinders their daily life, affects their health, and incurs a significant economic burden to society. Untreated Periodic Leg Movement (PLM) or Rapid Eye Movement Behaviour Disorder (RBD) could lead to a three to four-fold increased risk of stroke and Parkinson’s disease respectively. These risks bring about the need for less costly and more available diagnostic tools that will have great potential for detection and prevention. The goal of this study is to investigate the potentially clinically relevant but under-explored relationship of the sleep-related movement disorders of PLMs and RBD with cerebrovascular diseases. Our objective is to introduce a unique and efficient way of performing non-stationary signal analysis using sparse representation techniques. To fulfill this objective, at first, we develop a novel algorithm for Electromyogram (EMG) signals in sleep based on sparse representation, and we use a generalized method based on Leave-One-Out (LOO) to perform classification for small size datasets. In the second objective, due to the long-length of these EMG signals, the need for feature extraction algorithms that can localize to events of interest increases. To fulfill this objective, we propose to use the Non-Negative Matrix Factorization (NMF) algorithm by means of sparsity and dictionary learning. This allows us to represent a variety of EMG phenomena efficiently using a very compact set of spectrum bases. Yet EMG signals pose severe challenges in terms of the analysis and extraction of discriminant features. To achieve a balance between robustness and classification performance, we aim to exploit deep learning and study the discriminant features of the EMG signals by means of dictionary learning, kernels, and sparse representation for classification. The classification performances that were achieved for detection of RBD and PLM by means of implicating these properties were 90% and 97% respectively. The theoretical properties of the proposed approaches pertaining to pattern recognition and detection are examined in this dissertation. The multi-layer feature extraction provide strong and successful characterization and classification for the EMG non-stationary signals and the proposed sparse representation techniques facilitate the adaptation to EMG signal quantification in automating the identification process.