• Digital Humanities and Cultural Heritage
• Science and Innovation Policy Studi... close

This is a collection of data used for the article "Indicators for the use of Robotic Labs in Basic Biomedical Research". The primary result files are: - automated_method_dist.csv - non_automated_method_dist.csv - automated_methods_frequency.csv - non_ automated_methods_frequency.csv These were generated by running descriptive statistics of annotating the papers listed in articles_piis_dois.csv with the Medical Subject Headings 2015 vocabulary.

This report provides a documentation and detailed description of the data sources, structure, operationalisation and outputs of the European Social Innovation Database (ESID), a comprehensive and authoritative source of information on social innovation projects and actors in Europe and beyond. 

Artificial Neural Networks (ANNs) are being increasingly used in various disciplines outside computer science, such as bibliometrics, linguistics, and medicine. However, their uptake in the social science community has been relatively slow, because these highly non-linear models are difficult to interpret and cannot be used for hypothesis testing. Despite the existing limitations, this paper argues that the social science community can benefit from using ANNs in a number of ways, especially by outsourcing laborious data coding and pre-processing tasks to machines in the early stages of analysis. Using ANNs would enable small teams of researchers to process larger quantities of data and undertake more ambitious projects. In fact, the complexity of the pre-processing tasks that ANNs are able to perform mean that researchers could obtain rich and complex data typically associated with qualitative research at a large scale, allowing to combine the best from both qualitative and quantitative approaches.

Pastaruoju metu neuroninių tinklų modeliai yra vis dažniau taikomi įvairiose mokslo srityse (mokslometrijoje, lingvistikoje, medicinoje). Tačiau socialiniuose moksluose šie modeliai ir giliojo mokymosi metodai apskritai kelią skinasi sunkiai. Pagrindinės menko jų populiarumo priežastys yra tai, kad netiesinių modelių rezultatus yra sunku interpretuoti ir jie nėra tinkami hipotezėms tikrinti. Nepaisant to, šie metodai gali būti labai naudingi socialiniuose moksluose, ypač automatizuojant darbui labai imlias tyrimo proceso dalis. Straipsnyje pristatomas pavyzdys, kaip šie modeliai yra naudojami atliekant didelio masto teksto kodavimą. Pateikiamas pavyzdys rodo, kad kai kuriais atvejais šie metodai leidžia išvengti kompromiso tarp analizės gylio ir imties dydžio, taip pat leidžia kombinuoti geriausias kokybinių ir kiekybinių metodų praktikas. Artificial Neural Networks (ANNs) are being increasingly used in various disciplines outside computer science, such as bibliometrics, linguistics, and medicine. However, their uptake in the social science community has been relatively slow, because these highly non-linear models are difficult to interpret and cannot be used for hypothesis testing. Despite the existing limitations, this paper argues that the social science community can benefit from using ANNs in a number of ways, especially by outsourcing laborious data coding and pre-processing tasks to machines in the early stages of analysis. Using ANNs would enable small teams of researchers to process larger quantities of data and undertake more ambitious projects. In fact, the complexity of the pre-processing tasks that ANNs are able to perform mean that researchers could obtain rich and complex data typically associated with qualitative research at a large scale, allowing to combine the best from both qualitative and quantitative approaches.

Studies examining the similarity of online self-report survey responses using different item formats have yielded inconclusive results. Additionally, no studies have used appropriate methods for thoroughly and correctly examining equivalence across conditions. We examined the comparability of survey responses across four item formats—horizontal radio button, text box, drop-down menu, and vertical radio button—in two studies. The second study added two response categories: optional responding and forced responding. Participants were college students at two institutions of higher education who were randomly assigned to conditions. They completed measures of computer self-efficacy, personality, and social desirability. Results of both studies indicated quantitative (mean scores) and qualitative (internal consistency estimates and scale intercorrelations) equivalence. However, there were notable differences in auxiliary equivalence such that participants in the text box condition had lower amounts of missing data than those in the other conditions, those in the horizontal radio button condition completed the study in the shortest amount of time, and participants across conditions generally preferred to use drop-down menus compared to other item formats.

ObjectiveA vast amount of medical data is still stored in unstructured text documents. We present an automated method of information extraction from German unstructured clinical routine data from the cardiology domain enabling their usage in state-of-the-art data-driven deep learning projects.MethodsWe evaluated pre-trained language models to extract a set of 12 cardiovascular concepts in German discharge letters. We compared three bidirectional encoder representations from transformers pre-trained on different corpora and fine-tuned them on the task of cardiovascular concept extraction using 204 discharge letters manually annotated by cardiologists at the University Hospital Heidelberg. We compared our results with traditional machine learning methods based on a long short-term memory network and a conditional random field.ResultsOur best performing model, based on publicly available German pre-trained bidirectional encoder representations from the transformer model, achieved a token-wise micro-average F1-score of 86% and outperformed the baseline by at least 6%. Moreover, this approach achieved the best trade-off between precision (positive predictive value) and recall (sensitivity).ConclusionOur results show the applicability of state-of-the-art deep learning methods using pre-trained language models for the task of cardiovascular concept extraction using limited training data. This minimizes annotation efforts, which are currently the bottleneck of any application of data-driven deep learning projects in the clinical domain for German and many other European languages.

POSTDATA focused on poetry analysis, the publication of poetic resources and their exploration, applying Digital Humanities methods.   This is a trans-domain project, as it combines traditional philological studies with digital humanities technologies and tools. It is focused on poetry analysis (rhythm, accent, typology) analyzing and comparing databases of poetic repertories, analysing TEI-XML documents and databases that described poetic resources. The project is combining semantic web technologies to represent and publish the information as Linked Open Data in order to make the aforementioned resources interoperable, for a field, poetry, that has never been analyzed as a standardized and interoperable area.  POSTDATA is conceived as a digital humanities project that uses the richest poetry collections from 16 languages and literary traditions in combination with the most updated technologies: the construction of an ontology, Semantic Engineering Modelling, Natural Language Processing (NLP) and artificial intelligence, applying automation technologies and data mining to non-standard literary texts (i.e. Medieval Spanish)  Poetry Lab combines the application of Artificial Intelligence and NLP technologies with the creation of poetic corpora in different languages producing different tools and innovative applications, such as our geolocation tool for Spanish Medieval texts HisMeTag. On the other hand, the project has made great progress in the automatic detection of enjambment, a difficult prosodic phenomenon to analyse and understand even from the literary point of view.