s

Min Wei

Shandong University | Master

School of Computer Science

Interested direction

  • Conversational information extraction
  • Drug drug interaction prediction

Skilled language

LINUX PYTHON

Research experience

Conversational information extraction

2022.01 - 2022.06

  • This task receives a Chinese doctor-patient dialogue as input and predicts the item (symptoms, surgery, examination) and state (positive, negative, doctor positive, doctor negative) mentioned in the dialogue.
  • Implementation method: use pre-training language model (such as bart, mt5) to extract information by generating.

    • Drug drug interation prediction

    2022.06 -

  • This task receives two drugs as input and predicts whether their pharmacological effect is altered when coadministration.
  • Implementation method: Use GNN-based method to extract the substructures of two drugs, and predict their interaction through the interaction between the two drug substructures (functional groups).