Rashed Salem

An Ontology-Based Framework for Linking Heterogeneous Medical Data

Clinical records contain a massive heterogeneity number of data, generally written in free-note without a linguistic standard. Other forms of medical data include medical images with/without metadata (e.g., CT, MRI, radiology, etc.), audios (e.g., transcriptions, ultrasound), videos (e.g., surgery recording), and structured data (e.g., laboratory test results, age, year, weight, billing, etc.). Consequently, to retrieve the knowledge from these data is not trivial task. Handling the heterogeneity besides largeness and complexity of these data is a challenge. The main purpose of this paper is proposing a framework with two-fold. Firstly, it achieves a semantic-based integration approach, which resolves the heterogeneity issue during the integration process of healthcare data from various data sources. Secondly, it achieves a semantic-based medical retrieval approach with enhanced precision. Our experimental study on medical datasets demonstrates the significant accuracy and speedup of the proposed framework over existing approaches.

Automatic Rules Generation Approach for Data Cleaning in Medical Applications

Data quality is considered crucial challenge in emerging big data scenarios. Data mining techniques can be reutilized efficiently in data cleaning process. Recent studies have shown that databases are often suffered from inconsistent data issues, which ought to be resolved in the cleaning process. In this paper, we introduce an automated approach for dependably generating rules from databases themselves, in order to detect data inconsistency problems from large databases. The proposed approach employs confidence and lift measures with integrity constraints, in order to guarantee that generated rules are minimal, non-redundant and precise. The proposed approach is validated against several datasets from healthcare domain. We experimentally demonstrate that our approach outperform significant enhancement over existing approaches.

CDRT: An efficient clustering algorithm for distributed real-time database sites

Recently, the demand for real-time database is increasing. Most real-time systems are inherently distributed in nature. They need data to be obtained and updated in a timely fashion. Sometimes required data are at a particular location is not available, and needed to be obtained from remote site. This may take long time that make the temporal data invalid resulting in large number of tardy transactions with their fetal effect. Clustering the database sites nodes can help distributed real-time database systems to face the challenges meeting their time requirements. Reducing the large number of network sites into many clusters with smaller number of sites will effectively decrease the response time, resulting in better meeting of time constraints. In this work, we introduce a novel clustering algorithm for distributed real-time database that depend on both the communication time cost and the timing properties of data. The results showed lower communication time, higher database performance and better meeting of timing requirements.