DRUGLE - A semantic search engine for pharmacology and medical drug information

Welcome to the project site for DRUGLE. DRUGLE is a semantic search engine specialized for pharmacology and medical drug information. DRUGLE is a project sponsored by  Karolinska Institutet and managed by  Seibo Software Studios.

Intended audience and users

DRUGLE is meant to be a search engine for authorative information on pharmacology and medical drugs. As such it is intended to function both as a general point of reference as well as a decision support system.

The intended audience can be categorized in the following main user groups:

  • Healthcare personnel, ie doctors, nurses etc.
  • Patients and relatives of patients.
  • Researchers and academics.
  • Medical students.
  • Healthcare information & process managers.
  • Healthcare systems such as EHR and drug prescription systems.
  • The general user.

Furthermore, DRUGLE can easily be integrated with other applications and systems. Typical examples are EHR and drug prescription systems.

By continuously building and adapting a semantic knowledge base and by use of specialized search profiles DRUGLE will be able to serve the varying needs of different user groups in a consistent manner and via a standard interface.

Key Features

Notable key features of DRUGLE are:

  • Can hold large amounts of pharmacological and medical drug data.
  • Data gathered and continuously updated from a number of authoratitive sources on the web.
  • Gathered data is analyzed and structured so that the search resolution is higher than the individual documents themselves.
  • Easy to add and build indexes for new data sources.
  • Support for boolean and probabilistic search models.
  • Search algoritms is both model and terminology based.
  • Terminology solution supports synonyms and lexical proximity.
  • Profile and context based search features.
  • Adaptive semantic knowledge base.
  • Pharmacological and medical drug information model.
  • Programmatic REST API enabling web applications to easily integrate with DRUGLE.
  • Plugin architecture for extending web scraping features for new sources.

Notable key features of the DRUGLE web application are:

  • AJAX based web application.
  • Search for information in different languages.
  • Profile based search.

Design and implementation

DRUGLE is being designed and implemented in accordance with the following general principles:

One of the key principles underlying the design and implementation of DRUGLE is that data storage (persistence) and information retrieval are distinct yet complementary aspects of data management.

The following technologies are used: Python, Apache, Linux, Xapian, XHTML/CSS/Javascript, JSON, HTTP-based REST web services.

The following CM/QA tools and technologies are used: Subversion, SCons, Trac and the Debian package mangagement system.

The beta installation

The beta installation at  http://demo.drugle.info currently indexes the following information:

Note: SPC stands for Summary of Product Characteristics and is the formal specification required for registering drugs with national drug registration authorities.

Future plans and development

Watch this space!

KI and Seibo Software Studio are looking for partners who are interested in developing and/or contributing to DRUGLE as well as funding management and maintenance of the DRUGLE Beta installation.

Contact

For more information please contact one of:

  • Lars L Gustafsson, KI
  • Daniel Rodríguez, Seibo Software Studios

Copyright and terms of use

The DRUGLE beta installation is free to use. The source code used to implement DRUGLE is copyright of KI/Seibo Software Studios - © 2009. All rights reserved. The data used in the DRUGLE beta installation is freely available from the respective authorities.