A Semantic-Based Dynamic Search Engine Design and Implementation for Electronic Medical Records

A Semantic-Based Dynamic Search Engine Design and Implementation for Electronic Medical Records

Weider D. Yu (San Jose State University, USA) and Seshadri K. Yilayavilli (San Jose State University, USA)
Copyright: © 2012 |Pages: 13
DOI: 10.4018/978-1-4666-0909-9.ch023
OnDemand PDF Download:
No Current Special Offers


In the current technology driven world, information retrieval activities are in almost every aspect of daily, as society uses popular web search engines like Google, Yahoo!, Live Search, Ask, and so forth to obtain helpful information. Often, these popular search engines look for and obtain key information; however, not all of the retrieved items are relevant in context to the search target a. Thus, it is left for the user to filter out unwanted information, using only a few information items left from the search results. These popular web search engines use a first generation search service based on “static keywords”, which require the users to know exactly what they want to search and enter the right keywords. This approach puts the user at a disadvantage. In this paper, the authors investigate and design a dynamic, question-answer search engine that enables searching by attributes for more precise and relevant information in Electronic Medical Record (EMR) field.
Chapter Preview

The semantic-based search engine uses the Microsoft’s ASP.NET 2.0 Framework to create a web based system. This web system runs on Internet Information Services 5.1. AJAX (Asynchronous JavaScript and XML) extension is used for making this application more interactive and faster. The programs of the system are written in C# language.

The ASP.NET framework provides a rich set of controls. AJAX uses asynchronous data transfer between the browser and the server. It is faster and user interactive. The data source is in the form of a Resource Description Framework (RDF) (Resource Description Framework, 2007; RDFS, 2008). A .NET based RDF parser is used and RDF-XML instances are used as the data source. The RDF instances will be based on two ontologies which are defined as a part of this project. The domains used for creating the ontology are in the space which is currently lacking a structure and which could greatly benefit from having one defined. One of the domains chosen is health field which is a very critical area. It is the target field of the approach. The other field is animal kingdom, mainly for illustration or explanation purposes of the approach.

Currently, major software companies like Microsoft Inc. and Google Inc. are heavily invested in Universal Health Care. US Health Care is lacking a standardized way to organize and share patient health records. As a solution to this, a minimal ontology is created to define Electronic Medical Records (EMR). And the second ontology is for the Animal Kingdom which is important for educational reasons. Apart from the two mentioned, some existing ontology like Friend Of A Friend (FOAF) used for social networking is also used to demonstrate the power of this search engine. The instance of these ontologies forms the data source for data mining (Lacy, 2005).

A .NET based RDF parser is used to support mining of the information. Microsoft Web Services is used to retrieve these RDF instances from remote data stores or local data stores. It enables loose coupling and interoperability. The server provides deep XML integration and security enhancements giving tremendous flexibility.

The architecture for the semantic-based dynamic search engine web application follows a simple N-tier based architecture. This is as shown in Figure 1 below. The following diagram shows the various components and the flow of requests and data exchange between them.

Figure 1.

System architecture layers


Complete Chapter List

Search this Book: