NLP@Pitt

[Proposal Presentation] Heather Piwowar 01/27/2009 @ 12pm

venue: Senott Square Rm 6329
Many initiatives encourage research data sharing in hopes of increasing research efficiency and quality, but the effectiveness of these early initiatives is not well understood. The objective of this dissertation is to examine the feasibility of evaluating data sharing behavior based on examination of the biomedical literature.
The proposal has three aims:

Aim 1: Does sharing have benefit for those who share?
Aim 2: Can sharing and withholding be systematically measured?
Aim 3: How often is data shared? What predicts sharing?


Aim 2 involves NLP: First, I will use NLP on full-text biomedical research articles to identify those that generate gene expression microarray data. Second, to assess whether the authors of these data-generating studies share or withhold their data, I will investigate mining full text for statements of data submission and database citation fields for primary data references.

[ANNOTATION STUDY] by Danielle Mowry: Tuesday 12/16/08 $ 12pm

venue: Senott Square Rm 6329
In Biomedicine, simply indentifying conditions like symptoms, findings and disease in a report is not sufficient for many application areas including reimbursement coding, quality assurance and biosurveillance. Developing a reliable annotation scheme that captures contextual information about a given condition (whether it is present or absent, who is experiencing it, when was it experienced etc) is essential for distinguishing whether a statement about a condition is relevant for these purposes. We are currently developing an annotation schema that may be useful for these application areas. The purpose of this meeting is to conduct a pilot annotation study to obtain feedback from a wider audience about the application of the scheme and obtain suggestions for improvement. The meeting will proceed in three parts. First, a short 15 minute PowerPoint review of the scheme will be given. Then, subjects will be provided brief guidelines to reference and asked to annotate up to 10 sentences over 20 minutes. The session will conclude with an informal review of the annotations and participants will be encouraged to provide constructive feedback of the guidelines and annotation scheme. The meeting will take place between 12-1pm on Tuesday 12/16/08.