Joint Model for Entity Recognition and Normalization

Design a joint model utilizing neural network architecture for mention detection and entity linking within a biomedical ontology (i.e., UMLS concepts).

Document Retrieval using Neural Text Summarization

Scholarly document retrieval for precision medicine using deep learning methods for document classification and text summarization.

poster paper git
Misleading Information on Twitter Regarding Nicotine as Potential Prophylactic or Therapeutic for COVID-19

Predictive model using CBA algorithm (apriori association rules) for identifying tweets with misleading information, especially on the use of smoking (vaping) for COVID-19.

BMET (BioMedical Entity Tagged) Embeddings

Learning joint word-concept embeddings from the biomedical entity annotated corpus. We use the MeSH (Medical Subject Headings) codes for representing biomedical entities.

paper git
Identifying Special Interest Groups on Twitter and the Effects of Twitter's interventions

Data analytics on the perspectives of a special interest group (i.e., QAnon) towards the COVID related events.



Kratylos: Unified linguistic corpora from diverse data sources

developed a document management system that serves the (computational) linguistic researchers for managing and accessing the lexical and corpus datasets

Endangered Language Alliance (ELA) and University of Kentucky