Ontology-based deep learning for human behavior prediction with explanations in health social networks

Abstract

Human behavior modeling is a key component in application domains such as healthcare and social behavior research. In addition to accurate prediction, having the capacity to understand the roles of human behavior determinants and to provide explanations for the predicted behaviors is also important. Having this capacity increases trust in the systems and the likelihood that the systems actually will be adopted, thus driving engagement and loyalty. However, most prediction models do not provide explanations for the behaviors they predict. In this paper, we study the research problem, human behavior prediction with explanations, for healthcare intervention systems in health social networks. We propose an ontology-based deep learning model (ORBM+) for human behavior prediction over undirected and nodes-attributed graphs. We first propose a bottom-up algorithm to learn the user representation from health ontologies. Then the user representation is utilized to incorporate self-motivation, social influences, and environmental events together in a human behavior prediction model, which extends a well-known deep learning method, the Restricted Boltzmann Machine. ORBM+ not only predicts human behaviors accurately, but also, it generates explanations for each predicted behavior. Experiments conducted on both real and synthetic health social networks have shown the tremendous effectiveness of our approach compared with conventional methods.

A scoping review of ontologies related to human behaviour change

“…Techniques included annotating published literature during ontology development, that is, marking of documents to identify ontological terms 27,32 . Data from health social network datasets 38 and external devices 34 were also used in ontology development.…”

“…Descriptions of the identified ontologies are provided in Table 1, with a brief description of each ontology available in Supplementary Notes 1 and graphs depicting related measures provided in Supplementary Figure 1. Ontologies represented areas of mental processes and cognitions 25,26,27 , mental 28,29 and physical disease 30,31,32 , psychological experimental design 33 , emotions 34,35 , epidemiology 36 and healthcare 37,38 . Identified ontologies were typically of medium scale, featuring between 100-1000 classes (entities; Table 1), with only the OBO Foundry-approved Human Disease Ontology (DOID) 30 having >10,000 classes (entities).…”

Trajectory Prediction of Assembly Alignment of Columnar Precast Concrete Members with Deep Learning

“…They proposed a cnn-based technology that uses existing pedestrian detection technology to generate a sum difference framework, which is used as the input of CNN networks (such as AlexNet, GoogleNet and ResNet). In recent years, people have done a lot of research on data processing and prediction based on deep learning [15,21,22]. In order to gain the ability to make inferences from complex scenarios, as humans, Walker [23] proposed a simple visual prediction method, which combines intermediate visual elements with the validity of time modeling.…”

Ontology-based cloud computing approach implements on mobile healthcare and management

“…By using an interdisciplinary research approach, where user researchers, (ontology) engineers, researchers and domain stakeholders are at the forefront, a platform can be developed of great added value for the patients that want to grow old in their own home and for their caregivers [33]. Phan et al [34] proposed an ontology-based deep learning model (ORBM+) to predict human behavior over undirected and node-attributed graphs. A bottom-up algorithm was used to learn the user representation from health ontologies.…”

“…A bottom-up algorithm was used to learn the user representation from health ontologies. Sharma and Kaur [34] studied how effective social media has been proven to be in disseminating health information and to what extent methods to advance ontology have been helpful in extracting relevant knowledge from the large amount of healthcare data. Since ontology is a knowledge 7 engineering for the decomposition of knowledge in a specific problem domain, this study uses ontology to analyze the health care procedure, in order to develop an algorithm for cloud computing [35].…”