Abstract:

Temporal reasoning involves the deduction of temporal dependencies among temporal intervals, explanation of the past using historical and current data, planning and prediction of the future using temporal data. In this paper, we propose a diagnostic decision support system for adverse drug reaction using temporal reasoning. The major functionality of our system is focused on adverse drug reaction, which is an inadvertent medical consequence of treatment with pharmaceuticals. The analysis has been carried out based on Modified Association Classification algorithm, which uses Interestingness and Local Support measures to calculate the risk ratio and the odds ratio. We use a tool named Magnum Opus to generate rules, which are stored in a knowledge base. In addition to rule generation using Magnum Opus our system uses a modified version of Apriori algorithm to generate temporal rules. Given aquery, the system identifies the factors, which increases the risk of adverse drug reaction. We have built aknowledge base with an inference engine and a forecasting engine that applies the rules using a backward chaining control flow for effective prediction and decision-making.

Keywords: Drug Reaction, Knowledge Base, Temporal Reasoning, Prediction and Association Rules

(Full Paper 357KB)

BibTex:

@ARTICLE{P1120627001,

AUTHOR = {H.Khanna Nehemiah and A.Kannan},

TITLE = {A DIAGNOSTIC DECISION SUPPORT SYSTEM for ADVERSE DRUG REACTION USING TEMPORAL REASONING},

JOURNAL ={The International Journal of Artificial Intelligence and Machine Learning},

}

(Full Paper 357KB)