An Improved Differential Privacy-Preserving Truth Discovery approach In Healthcare

Nowadays, the occurrence of breast cancer amongst women has been increasing tremendously. To predict the breast cancer occurrence, physicians are in need of computer assisted way of feature extraction and classification approaches. In this scenario, there is even a greater need to maintain the privacy of patients personalized details. Many algorithms exist in literature that are used for data anonymization. The fundamental disadvantage of all these algorithms lies in limiting the ability to derive better value and insights from the data. In this proposed approach a differential privacy preserving algorithm is employed. This algorithm alters particulars of patients vital information in the data set. Thus, it not only improves the privacy of patient details compared to other existing approaches but also helps in getting accurate results. The advantage is that it will have the least impact on the accuracy of the truth discovery approach as the appropriate features for cancer prediction data set are not altered. For predicting the truth discovery, this proposed system employs Expectation- Maximization Algorithm. This proposed approach is first of its kind in combining the differential privacy preservation technique with truth discovery strategy. Finally, the performance evaluation was estimated in terms of accuracy, precision, recall, and F1-score and the outcomes are compared to that of the existing approaches to prove the effectiveness of the proposed mechanism.