Connecting to OpenAlex
Academic researchers often study a specific topic by looking for relevant papers in databases. However, this can be a challenge because some good papers might not appear in the search results due to limitations in the search criteria. For instance, if a paper is written in a language other than English or published in a journal that is not indexed by the database, it might be missed. On the other hand, searching too broadly can result in too many papers to review. To address this issue, we suggest using multi-language feature extraction techniques and training models on the OpenAlex database. This database is a comprehensive source of scientific knowledge that contains metadata for over 250 million works, including journal articles and books. By using this database, researchers can access a vast amount of relevant information and prepare for the next major advancement in academic research.
Progress
In the context of the FORAS project, a pipeline was created to search in the OpenAlex database using multilingual semantic vectors. This pipeline was used to search for relevant papers for a systematic review on PTSD trajectories. The search results were compared a number of different search strategies. See the links below for the data, scripts and preprint.
People involved
- Peter Lombaers - Lead
- Rens van de Schoot - Advisor