A Deep Learning Model for Mining Behavioral Preference of Home Care Demanders to Suppliers

EasyChair Preprint 8327

Abstract

This study aims at developing a deep learning prediction model for the preference of home care service demanders to suppliers available in a health care management platform. Firstly, a Node2vec-based algorithm enhanced by minimizing patients’ preference difference, is developed for capturing behavioural pattern of demanders; then an advanced deep learning GRU model, combing attention mechanism for service categories, is constructed to predict the visiting preference to various service suppliers by taking into account not only the constraints of technical requirement but also personal behavioural preference. Experimental results show that the proposed deep learning model, compared to the traditional deep learning GRU model, can significantly improve the quality of behavioural preference prediction, which can help improve the recommendation success rate of service suppliers to demanders.

Keyphrases: Attention Mechanism, deep learning, home care service, preference prediction

@booklet{EasyChair:8327,
  author    = {Hongying Fei and Mingzhu Xu},
  title     = {A Deep Learning Model for Mining Behavioral Preference of Home Care Demanders to Suppliers},
  howpublished = {EasyChair Preprint 8327},
  year      = {EasyChair, 2022}}