Canine Parvovirus Diagnosis Classification Utilizing Veterinary Free-Text Notes

Zhecheng Sheng, Emma Bollig, Jennifer Granick, Rui Zhang, Amanda Beaudoin

Abstract: This work demonstrates a NLP pipeline on classifying different canine parvovirus diagnosis from visit summaries. The preliminary results show promising efficacy in employing BERT based models for this task. This work also reveals a way to make use of the largely untapped data from veterinary free-text fields to help improve pet health.