The implementation of natural language processing and semantic technology in e-health website

Abstract

Searching for health information on the internet is the common activity nowadays. With the number of resources from the web increasing each day, the amount of available information is vast and with different range of context. This makes it more ambiguous and difficult if not challenging when trying to acquire specific information on specific health topics. Semantic Technology is the solution to resolve the problem of vagueness when retrieving information from the internet. It only provides high relevant results in accordance to the searched topic. On the other hand, Natural Language Processing (NLP) is a key point to success the semantic web. With implemented NLP, computer can analyse a given input in a human language and convert the input into a useful representation form. This research work provides a system platform that is based on the integration between the semantic technology and NLP where users can not only acquire useful information, but also, share the knowledge. The study presents a developed prototype that provides information related to the medical domain and specifically female related diseases. The female related disease under context includes cervical cancer as the main scope of this work. The web-based prototype is catered for both medical practitioners as well as public users. The system facilitates the knowledge sharing activities among medical practitioners and public users, as well as, allowing medical students in acquiring knowledge of typical enquiries on female related diseases made by patients. In addition, it serves as a tool for public awareness regarding protective and preventive measures against cervical cancer. The prototype includes ontology-like vocabulary database that is linked to the National Cancer Institute (NCI) thesaurus, and has been implemented and tested to a different range of respondents. The system provides an environment where the respondents can query anything related to the respective disease whether a medical jargon or a plain language within the same domain. It provides the most relevant answer to a posed query, as well as related terminologies and pictorial representation as aided information when necessary. The implemented and tested prototype has shown a significant and more precise response towards the users both in terms of time, and the context of the information being searched. The intended and involved respondents for which this system is designed, includes women with both medical background and nonmedical background. The study conducted for the semantic technology prototype for both respondents has shown user satisfaction in both categories.