Machine Learning is an increasingly important technology dealing with the growing complexity of the digitalised world. Despite the fact, that we live in a ‘Big data’ world where, almost ‘everything’ is digitally stored, there are many real-world situations, where researchers are still faced with small data samples. The present bibliometric knowledge synthesis study aims to answer the research question ‘What is the small data problem in machine learning and how it is solved?’ The analysis a positive trend in the number of research publications and substantial growth of the research community, indicating that the research field is reaching maturity. Most productive countries are China, United States and United Kingdom. Despite notable international cooperation, the regional concentration of research literature production in economically more developed countries was observed. Thematic analysis identified four research themes. The themes are concerned with to dimension reduction in complex big data analysis, data augmentation techniques in deep learning, data mining and statistical learning on small datasets.

Objective: The overall aim of the present study was to compare the coverage of existing research funding information for articles indexed in Scopus, Web of Science, and PubMed databases.Methods: The numbers of articles with funding information published in 2015 were identified in the three selected databases and compared using bibliometric analysis of a sample of twenty-eight prestigious medical journals.Results: Frequency analysis of the number of articles with funding information showed statistically significant differences between Scopus, Web of Science, and PubMed databases. The largest proportion of articles with funding information was found in Web of Science (29.0%), followed by PubMed (14.6%) and Scopus (7.7%).Conclusion: The results show that coverage of funding information differs significantly among Scopus, Web of Science, and PubMed databases in a sample of the same medical journals. Moreover, we found that, currently, funding data in PubMed is more difficult to obtain and analyze compared with that in the other two databases. This article has been approved for the Medical Library Association’s Independent Reading Program.

Background: Children with developmental disabilities may need support with motor skills such as balance improvement, cognitive skills such as vocabulary learning, or social skills such as adequate interpretation of emotional expressions. Digital interactive games could support the standard treatments. We aimed to review clinical studies which investigated the application of serious games in children with developmental disabilities. Methods: We searched MEDLINE and Scopus on 05 May 2019 limited to the English language. We included people between two and 24 years of age who were affected by neurodevelopmental disorders and who received digital serious game-based medical interventions such as any computer- based or video-based games. We considered any study design reporting primary data. We used title, abstract, and full-text of journal articles to build diagnostic groups, and we described some selected specific game applications. Results: The majority of the 145 relevant studies reported on autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), developmental coordination disorder (DCD), and disabilities affecting intellectual abilities (DAIA). 30 of the 145 studies reported a randomized design. We detailed six specific applications aimed at improving abilities in children with ASD, ADHD, cerebral palsy, and Down syndrome. We visualized the diagnostic groups by bibliographic mapping, and limited the text to the title and abstract of journal articles. Conclusion: We identified promising results regarding anxiety reduction, stress regulation, emotion recognition, and rehabilitation. Currently, there appears to be a lack of clinical evidence that children with neurodevelopmental disorders can benefit from the application of serious games.

In response to the digital era revolution occurring toward the end of the millennium, libraries had to change their business models in order to deliver more advanced and smart services. Meanwhile, the knowledge doubling process encountered in the last century caused the exponential increase in the number of research publications. Consequently, it is impossible to read all the new published articles manually, and even harder to predict new research directions. Thus, identifying prospective research topics could be one of the smart services provided by the digital-era hospital library. A case of such service was implemented in this study, using bibliometric mapping. Two corpora concerning medical informatics harvested from PubMed database (one for the period 2014/15 and the other for 2016/17) were analyzed and compared. Based on the comparison possible hot future research topics were identified.