A basket with the latest published research in communication science. Every day at 2 AM UTC, I query the Crossref API for new research articles that appeared in the previous 7 days across 110 journals in communication science and adjacent fields. Additionally, I search major preprint servers for new papers that may be of interest to communication science researchers.
Keeping up with newly published research in communication science can be challenging. Email alerts from publishers clutter inboxes and arrive at irregular intervals. Setting up alerts or finding RSS feeds for multiple publishers is time-consuming. Social media platforms like Twitter have their own limitations for academic purposes.
Inspired by Moritz Marbach’s Paper Picnic project for Political Science, this adaptation of his project focuses on communication science research with some love for politics and related social science areas that are helpful to my research.
All journal data is sourced from the Crossref Crossref API. Preprints are retrieved from the Open Science Foundation API.
The backend consists of a crawler written in R, hosted in a GitHub repository. Every morning, GitHub Actions executes the crawler. The resulting data is stored in a JSON file and rendered into an HTML file using GitHub Pages.
For each journal, the crawler retrieves articles added in the previous week. It requests all articles with “created” or “published” fields in the Crossref database within the last seven days.
The crawler collects title, authors, full-text link, and abstract information. However, some publishers, like Elsevier or Taylor & Francis, do not include abstracts in their Crossref metadata (see this Crossref Blog for details). Unfortunately, a large portion of communication journals are partnered with publishers who do not allow any automated retrieval of their abstracts.
As journals typically have two ISSN numbers (print and electronic, see here), the crawler retrieves articles for both and deduplicates the results (usually). ISSN numbers are obtained from the Crossref lookup tool.
For generic titles (e.g., “Errata”, “Frontmatter”, “Backmatter”), the crawler adds a filter tag. For articles from multidisciplinary journals and preprint, GPT-4 is prompted to determine if the content is relevant to communication science (intending to err on the side of inclusiveness). Filtered content is hidden by default but can be displayed by clicking the +N button at the top left of each journal section.
Find and fix bugs or add new features to the crawler/web page. Help out Moritz on his original version or make suggestions in mine.
Use the crawled data for your own tool:
Build an improved (and equally open source) version of this page.
Support The Initiative for Open Abstracts.