SMAFIRA - Artificial Intelligence for Finding Alternative Methods
https://smafira.bf3r.de/
Animal experiments and alternative methods
Before an animal experiment can be conducted as part of a scientific project, it needs to be approved by a competent authority. For this purpose, a researcher has to submit an application for approval in which the fulfillment of scientific and legal requirements is outlined. Part of this application is a thorough literature search to ensure that the planned animal experiment cannot be replaced by an alternative method. Such an alternative would be, for example, a method or procedure that is suitable to answer a specific scientific question without the use of live (vertebrate) animals, e.g., in vitro procedures. The search for possible alternative methods is often very complex and is a major challenge for researchers.
SMAFIRA – Artificial intelligence supporting the search for alternative methods
Although there are already search engines and literature databases for biomedical questions, which also provide semantic techniques, there is still no satisfying solution for the search for alternative methods to animal experiments. To support researchers, the Bf3R is developing a search engine for alternative methods to animal experiments that is based on the freely accessible biomedical literature database PubMed (Medline) and that enhances this database with important functions. SMAFIRA is an acronym for SMArt Feature based Interactive RAnking and shall enable scientists to find suitable suggestions for alternative methods to a given animal experiment (= reference document). Furthermore, SMAFIRA will rank the results of the search, i.e. the reference list with respect to their thematic correspondence to the given reference document and their relevance as a potential alternative method to the respective animal experiment. SMAFIRA incorporates state-of-the-art methods of text mining (e.g. Information Retrieval, Named-Entity Recognition or Relation Extraction) and machine learning (e.g. Neural Networks).
The first version of the searching engine SMAFIRA is available here.
Literature
- Neves M, Klippert A, Knöspel F, Rudeck J, Stolz A, Ban Z, Becker M, Diederich K, Grune B, Kahnau P, Ohnesorge N, Pucher J, Schönfelder G, Bert B, Butzke D Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments (2023). Journal of Biomedical Semantics:14(13).[under review, https://doi.org/10.1186/s13326-023-00292-w]
- Neves M. Integration of the PubAnnotation ecosystem in the development of a web-based search tool for alternative methods, Genomics & Informatics, 2020, 18(2). [full text and pdf, https://doi.org/10.5808/gi.2020.18.2.e18]
- Butzke D, Dulisch N, Dunst S, Steinfath M, Neves M, Mathiak B, Grune B. SMAFIRA-c: A benchmark text corpus for evaluation of approaches to relevance ranking and knowledge discovery in the biomedical domain(2020) Research Square. Preprint from Research Square. [full text and pdf, https://doi.org/10.21203/rs.3.rs-16454/v1]
- Neves M, Butzke D, Grune B. Evaluation of Scientific Elements for Text Similarity in Biomedical Publications(2019) 6th Workshop on Argument Mining,Association for Computer Linguistics. [pdf and bibtex, http://dx.doi.org/10.18653/v1/W19-4515]
- Neves M, Butzke D, Schönfelder G, Grune B. Bf3R at SemEval-2018 Task 7: Evaluating Two Relation Extraction Tools for Finding Semantic Relations in Biomedical Abstracts (2018), Proceedings of the 12th International Workshop on Semantic Evaluation, Association for Computational Linguistics:816-820. [pdf and bibtex, http://dx.doi.org/10.18653/v1/S18-1130]
- Neves M, Ševa J. An extensive review of tools for manual annotation of documents, (2019) Briefings in Bioinformatics: 22(1) 146-163 [full text, https://doi.org/10.1093/bib/bbz130]