An efficient reachability query based pruning algorithm in e-health scenario

doi:10.1016/j.jbi.2019.103171

Return

An efficient reachability query based pruning algorithm in e-health scenario

2019-06-01

PRE

Mondal, Fikureshi *

Nandini Mukherjee

DOI：10.1016/j.jbi.2019.103171

Original

Original request for help

Save

Abstract

En 中文

We propose a Disease - Symptom graph database for our mobile-assisted e-healthcare application. A large Disease - Symptom graph is stored in the cloud and accessed using mobile devices over the Internet. Query and search are the fundamental operations of graph databases. However, while searching the Disease - Symptom graph for making preliminary diagnosis of diseases, queries become complex due to the complex structure of data and also queries are too hard to write and interpret. Moreover, it is not possible to access the graph frequently due to limited bandwidth of the network, transmission delay, and higher cost. Subgraph generation or pruning algorithm for appropriate inputs is one of the solutions to this problem. In this paper, we propose an efficient pruning algorithm by introducing a new approach to decompose the Disease - Symptom graph into a series of symptom trees (ST). All the Symptom trees are merged to build a pruned subgraph which is our requirement. We demonstrate the efficiency and effectiveness of our pruning algorithm both analytically and empirically and validate on Disease - Symptom graph database, as well as other real graph databases. Also a comparison is done with an efficient existing reachability based Chain Cover algorithm after modifying it ChainCoverPrune as pruning algorithm. These two algorithms are tested for storage and access parametric measures for querying the synthetic and real directed databases to show the efficiency of the proposed algorithm.

Keywords:

Disease - Symptom graph database

Reachability Query

Chain Cover

Pruning algorithm

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.