JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]


Journal of Information Science and Engineering, Vol. 36 No. 2, pp. 365-376


Soft Computing Model to Predict Chronic Diseases


THEYAZN H. H. ALDHYANI1, ALI SALEH ALI ALSHEBAMI1 AND MOHAMMED YAHYA ALZAHRANI2
1Community College of Abqaiq
King Faisal University
Ahsaa, 31982 Kingdom Saudi Arabia
E-mail: {taldhyani; aalshebami}@kfu.edu.sa

2Information Technology Department
Albaha University
Albaha, 65527 Kingdom Saudi Arabia
E-mail: msawileh@bu.edu.sa


The World Health Organization (WHO) has reported that non-communicable diseases are too risky as one of the serious diseases that threaten this world. The chronic diseases are extremely complex, so collaboration with ecological, biological and behavioral has given challenges for researchers and developers to predict non-communicable diseases. Digital surveillance system aimed to detect search queries that help to improve the awareness and timeline of predicted health outbreak. The purpose of the research study is to use Google Trend data for predicting chronic diseases. In this study, soft computing algorithm is applied to map the web search activity behavior of the population to prevent chronic disease risk factors. The Google Trend search activity is used to identify relevant web search activity, so the study period was taken from the first week of January 2017 to last week of December 2017 as it consisted of four chronic diseases namely asthma, heart, diabetics, and kidney. The clinical data of chronic diseases were collected from the Centers for Disease Control and Prevention (CDC) to test and evaluate the proposed model. The Adaptive Neuro-Fuzzy Inference System (ANFIS) algorithm was used to map web search activity from Google Trend using CDC clinical data. The consideration data in 2017 has shown there is strongly correlation between web search activity and clinical data. The standard evaluation metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE) were applied to evaluate the performance of the proposed model. The experimental results have shown that there is a relationship between internet search queries and clinical data, and thus prediction errors are very less. The high predictive validity of web search queries for chronic diseases have given the possibility to consider the population web information in order to predict Non-Communicable Disease (NCD) risk for avoiding and spreading in a large area scale. It is concluded that the proposed system can help to detect and predict the non-communicable diseases in the earliest stage.


Keywords: chronic diseases, soft computing, Google trend data, intelligent model, machine intelligent

  Retrieve PDF document (JISE_202002_14.pdf)