[Asis-l] Researcher Position (IR, LM, DM/TM, NLP) at University of Washington, Information School & VA

[Efthimis N. Efthimiadis]

Researcher in information retrieval, language modeling, text classification and mining, and NLP.

Information School, University of Washington, Seattle, WA, USA

Duration: appointment 12-15 months.
Application Deadline: June 30, 2008
Location:  Information School, UW, and VA HSR&D offices, both in Seattle, WA, USA


Position Description:

We are seeking a Researcher (with background in IR, LM, DM/TM or NLP) for a project that involves the study of Electronic Medical Records.  Please review the information below and don't hesitate to contact Efthimis Efthimiadis (efthimis at u.washington.edu) or Ken Hammond (Kenric.Hammond at va.gov) for additional information.


TITLE OF PROJECT:  Assessing Information Value in Computerized Patient Care Documentation Systems

1. BACKGROUND AND PURPOSE OF RESEARCH.

Computerized patient care documents (CPD), consisting of progress notes, summaries, and reports documenting medical care are a key part of modern computerized health care information systems. On-line documents are more accessible and legible than paper records, but the transition to computer systems has introduced new challenges. Computerized documents differ from traditional documents: document writers may copy text from other documents; ancillary data is frequently inserted automatically; and templates assisting text entry introduce boilerplate text. Shortcuts aid writing documents but often result in poor readability and reduce the value of on-line documentation. Copying can propagate errors and obscure the original source of information. Data overload is a significant problem: Nationally, Veterans Administration (VA) systems currently store over 700 million patient documents. Individual charts may contain hundreds of documents. Computerized medical record users can be hard pressed to find the specific information they need amid a flood of potentially useful information, because the time available to retrieve information from documentation is constrained.

Improved patient records will benefit patients, practitioners, and anyone else who uses them. The objective of this part of the research project is to analyze and characterize a large dataset of medical records using information retrieval, language modeling, text classification and mining, and natural language approaches.


2. DATA SET

The VISN 20 Dataset, consists of approximately 18 million documents belonging to about 80,000 veterans, and has been extracted from computerized records kept at the VISN 20 data center (Vancouver, WA). The VISN 20 dataset is available as a SQL Server relational database, and it is located in a secure storage network at the HSR&D offices in Seattle.


CHARACTERISTIC DUTIES AND RESPONSIBILITIES

Required Qualifications
*       Experience with one or more of the following fields:
        -- Information Retrieval,
        -- Natural Language Processing,
        -- Language Modeling,
        -- Text Mining,
        -- Machine Learning
*       Familiarity with text retrieval software (e.g., Lucene)
*       Database management using SQL server in a Windows environment

*       A degree in medical informatics, information science, computer science,
        information systems or a related area
        (We welcome applicants about to finish their thesis).

*       Experience in managing and supervising software projects;
*       Excellent programming skills, preferably in Java;
*       Excellent communication and organization skills
*       Self-motivated with the ability to take initiative and work independently
*       Ambition to produce internationally recognized research output, and
*       Willingness to collaborate within an interdisciplinary team.

Successful Candidates will be required to:
*       Complete the UW Human Subjects training
                (http://www.washington.edu/research/hsd/training.html)
*       Adhere to HIPAA regulations
                (http://www.hhs.gov/ocr/hipaa)


For more information please contact:

Efthimis Efthimiadis (efthimis at u.washington.edu) or
Ken Hammond (Kenric.Hammond at va.gov)


To apply:

Apply online at:
 https://uwhires.admin.washington.edu/eng/candidates/default.cfm?szLocationID=88

Search for requisition - 44457


* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
Efthimis N. Efthimiadis <efthimis at u.washington.edu>
Associate Professor
The Information School, University of Washington
Suite 370 Mary Gates Hall, Box 352840
Seattle, WA 98195-2840, USA
Tel: (off.) 206-616-6077, (sch) 206-685-9937, fax. 206-616-3152 http://faculty.washington.edu/efthimis
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *