An introduction of the Hybrid Publications group of the research project Going Hybrid. The blog-post discusses affiliated researchers and organizations, working questions, expert opinions, and previous work done by partners.
LINK
Objective: To annotate a corpus of randomized controlled trial (RCT) publications with the checklist items of CONSORT reporting guidelines and using the corpus to develop text mining methods for RCT appraisal. Methods: We annotated a corpus of 50 RCT articles at the sentence level using 37 fine-grained CONSORT checklist items. A subset (31 articles) was double-annotated and adjudicated, while 19 were annotated by a single annotator and reconciled by another. We calculated inter-annotator agreement at the article and section level using MASI (Measuring Agreement on Set-Valued Items) and at the CONSORT item level using Krippendorff's α. We experimented with two rule-based methods (phrase-based and section header-based) and two supervised learning approaches (support vector machine and BioBERT-based neural network classifiers), for recognizing 17 methodology-related items in the RCT Methods sections. Results: We created CONSORT-TM consisting of 10,709 sentences, 4,845 (45%) of which were annotated with 5,246 labels. A median of 28 CONSORT items (out of possible 37) were annotated per article. Agreement was moderate at the article and section levels (average MASI: 0.60 and 0.64, respectively). Agreement varied considerably among individual checklist items (Krippendorff's α= 0.06–0.96). The model based on BioBERT performed best overall for recognizing methodology-related items (micro-precision: 0.82, micro-recall: 0.63, micro-F1: 0.71). Combining models using majority vote and label aggregation further improved precision and recall, respectively. Conclusion: Our annotated corpus, CONSORT-TM, contains more fine-grained information than earlier RCT corpora. Low frequency of some CONSORT items made it difficult to train effective text mining models to recognize them. For the items commonly reported, CONSORT-TM can serve as a testbed for text mining methods that assess RCT transparency, rigor, and reliability, and support methods for peer review and authoring assistance. Minor modifications to the annotation scheme and a larger corpus could facilitate improved text mining models. CONSORT-TM is publicly available at https://github.com/kilicogluh/CONSORT-TM.
DOCUMENT
Objective:Acknowledging study limitations in a scientific publication is a crucial element in scientific transparency and progress. However, limitation reporting is often inadequate. Natural language processing (NLP) methods could support automated reporting checks, improving research transparency. In this study, our objective was to develop a dataset and NLP methods to detect and categorize self-acknowledged limitations (e.g., sample size, blinding) reported in randomized controlled trial (RCT) publications.Methods:We created a data model of limitation types in RCT studies and annotated a corpus of 200 full-text RCT publications using this data model. We fine-tuned BERT-based sentence classification models to recognize the limitation sentences and their types. To address the small size of the annotated corpus, we experimented with data augmentation approaches, including Easy Data Augmentation (EDA) and Prompt-Based Data Augmentation (PromDA). We applied the best-performing model to a set of about 12K RCT publications to characterize self-acknowledged limitations at larger scale.Results:Our data model consists of 15 categories and 24 sub-categories (e.g., Population and its sub-category DiagnosticCriteria). We annotated 1090 instances of limitation types in 952 sentences (4.8 limitation sentences and 5.5 limitation types per article). A fine-tuned PubMedBERT model for limitation sentence classification improved upon our earlier model by about 1.5 absolute percentage points in F1 score (0.821 vs. 0.8) with statistical significance (). Our best-performing limitation type classification model, PubMedBERT fine-tuning with PromDA (Output View), achieved an F1 score of 0.7, improving upon the vanilla PubMedBERT model by 2.7 percentage points, with statistical significance ().Conclusion:The model could support automated screening tools which can be used by journals to draw the authors’ attention to reporting issues. Automatic extraction of limitations from RCT publications could benefit peer review and evidence synthesis, and support advanced methods to search and aggregate the evidence from the clinical trial literature.
MULTIFILE
All Gone is a series of experiments with AI that build on existing collections of climate fiction to create much-needed new climate imaginaries. As the climate crisis is also a “crisis of imagination” (Ghosh, 2016), this project turns to the art genre that is best at forecasting and imagining alternative futures: science fiction. Using collections of ‘cli-fi’ novels, in which science fiction meets natural disaster or heavy weather, algorithms are trained until they are able to render new climate imaginaries in textual and visual form. The edited texts and curated images are further developed into audio stories and a tarot deck as tools for reflection on present and future living with a changing climate.
MULTIFILE
In recent years, the number of publications on innovation in the construction industry has increased. Many of these documents address qualitative issues, e.g. policies for innovation and present case studies. A more quantitative approach is taken in this paper, which is the continuation of a previous study. It focuses on main types and sources of innovation in the construction industry, and includes an analysis of 55 years of publications in two leading Dutch professional journals. The results show a recent increase in innovation, with two-thirds of innovations coming out of supplying industries. Construction companies contribute mainly in process innovations. Innovation in construction remains to be technology- rather than market-driven. Regulations have a surprising impact, as over one-third of all counted new innovations are related to new regulations.
DOCUMENT
The definition of ‘Assistive Technology’ (AT) includes both assistive products and the services or actions necessary for safe and effective provision of the assistive products to people who need them. International standards and product specifications exist for assistive products. Despite huge unmet need for effective AT provision, a variety of service delivery models across different countries, and a shortage of personnel trained in this field, nowidely useable and accepted AT service provision guidelines currently exist. Aligned with contemporary global initiatives to improve access to AT, a scoping review was commissioned to inform the development of globally useable provision guidance. The aim was to deliver a rapid scoping review of the literature regarding quality guidelines for AT service provision. Method: The rapid scoping review utilised a two-tiered approach to identifying relevant publications: 1) systematic search of academic databases; 2) consultation with assistive technology organisations. The review was conducted in March 2023 across four databases (Medline, CINAHL, SCOPUS and Google Scholar) with no date limitations. Systematic outreach to international and global AT networks was used to access expert informants. Non-English publications were included utilizing Google Translate and support from expert informants to verify content. Analysis was guided by the body of work on quality AT provision and service delivery processes in Europe, as well as the World Health Organization-GATE 5P framework for strengthening access to AT. Results: The search strategies yielded 41 publications from diverse countries, and directed at differing assistive products, personnel and provision contexts. Results are reported from the charted data through to the data extraction framework, including type of publication, study design, audience and reach. We report on the type of AT and the AT provision ecosystem elements discussed, and service delivery process or steps and quality criteria service delivery. Conclusion: This review did not find established guidelines or standards for service provision, but it did identify key service delivery steps which may form part of such guidelines, and many of the 3 publications included mentioned the need for practice guidelines. Despite different contexts such as type of assistive product, recipient of the guidance, language, location and authorship, core elements of AT provision including service delivery steps can be identified. Consideration regarding the nuances of vocabulary, of process, and of enabling flexible foci, is recommended in systematizing globally applicable guidance. This review offers a strong starting point for developing guidance for assistive technology provision to meet global need.
DOCUMENT
Background: The combination of self-tracking and persuasive eCoaching in automated interventions is a new and promising approach for healthy lifestyle management. Objective: The aim of this study was to identify key components of self-tracking and persuasive eCoaching in automated healthy lifestyle interventions that contribute to their effectiveness on health outcomes, usability, and adherence. A secondary aim was to identify the way in which these key components should be designed to contribute to improved health outcomes, usability, and adherence. Methods: The scoping review methodology proposed by Arskey and O'Malley was applied. Scopus, EMBASE, PsycINFO, and PubMed were searched for publications dated from January 1, 2013 to January 31, 2016 that included (1) self-tracking, (2) persuasive eCoaching, and (3) healthy lifestyle intervention. Results: The search resulted in 32 publications, 17 of which provided results regarding the effect on health outcomes, 27 of which provided results regarding usability, and 13 of which provided results regarding adherence. Among the 32 publications, 27 described an intervention. The most commonly applied persuasive eCoaching components in the described interventions were personalization (n=24), suggestion (n=19), goal-setting (n=17), simulation (n=17), and reminders (n=15). As for self-tracking components, most interventions utilized an accelerometer to measure steps (n=11). Furthermore, the medium through which the user could access the intervention was usually a mobile phone (n=10). The following key components and their specific design seem to influence both health outcomes and usability in a positive way: reduction by setting short-term goals to eventually reach long-term goals, personalization of goals, praise messages, reminders to input self-tracking data into the technology, use of validity-tested devices, integration of self-tracking and persuasive eCoaching, and provision of face-to-face instructions during implementation. In addition, health outcomes or usability were not negatively affected when more effort was requested from participants to input data into the technology. The data extracted from the included publications provided limited ability to identify key components for adherence. However, one key component was identified for both usability and adherence, namely the provision of personalized content. Conclusions: This scoping review provides a first overview of the key components in automated healthy lifestyle interventions combining self-tracking and persuasive eCoaching that can be utilized during the development of such interventions. Future studies should focus on the identification of key components for effects on adherence, as adherence is a prerequisite for an intervention to be effective.
DOCUMENT
DOCUMENT
