Update README.md

README.md

@@ -16,4 +16,20 @@ task_categories:
 
 This can be used to finetune a decoder model for Q/A interaction, 
 alternatively it can be used to create *(question, positive, negative)* 
-triplets to train a sentence encoder using [SBERT](https://www.sbert.net/).
+triplets to train a sentence encoder using [SBERT](https://www.sbert.net/).
+
+Reference:
+
+'''
+@inbook{Nentidis_2023,
+   title={Overview of BioASQ 2023: The Eleventh BioASQ Challenge on Large-Scale Biomedical Semantic Indexing and Question Answering},
+   ISBN={9783031424489},
+   ISSN={1611-3349},
+   url={http://dx.doi.org/10.1007/978-3-031-42448-9_19},
+   DOI={10.1007/978-3-031-42448-9_19},
+   booktitle={Experimental IR Meets Multilinguality, Multimodality, and Interaction},
+   publisher={Springer Nature Switzerland},
+   author={Nentidis, Anastasios and Katsimpras, Georgios and Krithara, Anastasia and Lima López, Salvador and Farré-Maduell, Eulália and Gasco, Luis and Krallinger, Martin and Paliouras, Georgios},
+   year={2023},
+   pages={227–250} }
+'''