Search

Accurately identifying different representations of the same real–world entity is an integral part of data cleaning and many methods have been proposed to accomplish it. The challenges of this entity resolution task that demand so much research attention are often rooted in the task–specificity and user–dependence of the process. Adopting deep learning techniques has the potential to lessen these challenges. In this paper, we set out to devise an entity resolution method that builds on the robustness conferred by deep autoencoders to reduce human–involvement costs. Specifically, we reduce the cost of training deep entity resolution models by performing unsupervised representation learning. This unveils a transferability property of the resulting model that can further reduce the cost of applying the approach to new datasets by means of transfer learning. Finally, we reduce the cost of labeling training data through an active learning approach that builds on the properties conferred by the use of deep autoencoders. Empirical evaluation confirms the accomplishment of our cost–reduction desideratum, while achieving comparable effectiveness with state–of–the–art alternatives.

Multi-Label Text Classification (MLTC) is a well-known NLP task that allows the classification of texts into multiple categories indicating their most relevant domains. However, training model tasks on texts from web user deal with redundancy or ambiguity of linguistic information. In this work, we propose a comparative study about different neural network models for a multi-label text categorisation task in finance domain. Our main contribution consists of presenting a new annotated dataset that contains ∼26k posts from users associated to finance categories. To build that dataset, we defined 10 specific-domain categories that cover financial texts. To serve as a baseline, we present a comparative study analysing both the performance and training time of different learning models for the task of multilabel text categorisation on the new dataset. The results show that transformer-based language models outperformed RNN-based neural networks in all scenarios in terms of precision. However, transformers took much more time than RNN models to train an epoch model.

Healthcare experts have recently turned towards the use of Biosimulation models to understand the multiple or different causative factors that cause impairment in human organs. The applications of biosimulations have been applied in different biological systems ranging from human ear, cardiovascular, to neurovascular research using Finite Element Method (FEM). FEM provide a mathematical framework to simulate these dynamic biological systems. Visualizing and analyzing huge amounts of Finite Element (FE) Biosimulation numerical data is a strenuous task. In this paper, we demonstrate a Linked Data Visualiser -- called SIFEM Visualiser -- to help domain-experts to Visualise, analyze and compare biosimulation results from heterogeneous, complex, and high volume numerical data. The SIFEM Visualiser aims to help healthcare experts in exploring and analyzing Finite Element (FE) numerical data and simulation results obtained from different aspects of inner-ear (Cochlear) model - such as biological, geometrical, mathematical, and physical models. This publication has emanated from research supported in part by the research grant from Science Foundation Ireland (SFI) under Grant Number SFI/12/RC/2289 and EU project SIFEM (contract Number 600933). REFERE non-peer-reviewed

Biosimulation researchers use a variety of models, tools and languages for capturing and processing different aspects of biological processes. However, current modeling methods do not capture the underlying semantics of the biosimulation models sufficiently to support building, reusing, composing and merging complex biosimulation models originating from diverse experiments. In this paper, we propose an ontology based and multi-layered biosimulation model to facilitate researchers to share, integrate and collaborate their knowledge bases at Web scale. In particular, we investigate the semantic biosimulation model under the context of the multi-scale finite element (FE) modelling of the inner-ear. The proposed ontology-based biosimulation model will provide a homogenized and standardized access to the shared, semantically integrated and harmonized datasets for clinical data (histological data, micro-CT images of the cochlea, pathological data) and inner ear FE simulation models. The work presented in this paper is analyzed and designed as part of the SIFEM EU project.

The integration of a small fraction of the information present in the Web of Documents to the Linked Data Web can provide a significant shift on the amount of information available to data consumers. However, information extracted from text does not easily fit into the usually highly normalized structure of ontology-based datasets. While the representation of structured data assumes a high level of regularity, relatively simple and consistent conceptual models, the representation of information extracted from texts need to take into account large terminological variation, complex contextual/dependency patterns, and fuzzy or conflicting semantics. This work focuses on bridging the gap between structured and unstructured data, proposing the representation of text as structured discourse graphs (SDGs), targeting an RDF representation of unstructured data. The representation focuses on a semantic best-effort information extraction scenario, where information from text is extracted under a pay-as-you-go data quality perspective, trading terminological normalization for domain-independency, context capture, wider representation scope and maximization of textual information capture.

The vision of creating a Linked Data Web brings together the challenge of allowing queries across highly heterogeneous and distributed datasets. In order to query Linked Data on the Web today, end-users need to be aware of which datasets potentially contain the data and also which data model describes these datasets. The process of allowing users to expressively query relationships in RDF while abstracting them from the underlying data model represents a fundamental problem for Web-scale Linked Data consumption. This article introduces a multidimensional semantic space model which enables data model independent natural language queries over RDF data. The center of the approach relies on the use of a distributional semantic model to address the level of semantic interpretation demanded to build the data model independent approach. The final multidimensional semantic space proved to be flexible and precise under real-world query conditions achieving mean reciprocal rank = 0.516, avg. precision = 0.482 and avg. recall = 0.491.

In this work, we implemented a brain-machine interface (BMI) based on electroencephalographic (EEG) signals and used it to classify and separate three types of mental tasks: motor imagery with the right and left hands and simple arithmetic sums. In order to reduce dimension of variables and increase classification power, we used both PCA and ICA based algorithms for spectral analysis. Our results show that we were no able to reduce dimension without reducing classification performance.