Title: Designing Adaptive Feedback for Improving Data Entry Accuracy
Author: Kuang Chen, Joseph M. Hellerstein
Publisher: UIST' 10, October 3-6, 2010 New York
Summary
The authors of this article developed USHER, a theoretical data-driven foundation used to improve the quality of data entry. It can be difficult to spot errors in data entry until it's too late. USHER, based on prior data, learns a probabilistic model of the dependencies between form questions and values.
It uses this information learned to maximize information gain. It begins by asking unpredictable questions first and is therefore able to better predict answers for remaining questions. In this paper, the authors use USHER's predictive ability to improve data entry accuracy and efficiency.
They applied their mechanisms to professional data entry clerks using real patient data in 6 clinics in Uganda. Their results showed that they were able to reduce potential error by up to 78%. They believe that their approach has a wide range of applicability and can be very important for decision-making and resource allocation.
Discussion
I thought the goal of the researchers was a very valid one. Data entry is something that people do all the time but usually make quite a few errors. I always double and even triple-check when I perform data entry as errors have caused many problems for myself. Something like this has a large range of applicability, something that I applaud them for because it's been a while since i've read a paper I have felt that way.
It uses this information learned to maximize information gain. It begins by asking unpredictable questions first and is therefore able to better predict answers for remaining questions. In this paper, the authors use USHER's predictive ability to improve data entry accuracy and efficiency.
They applied their mechanisms to professional data entry clerks using real patient data in 6 clinics in Uganda. Their results showed that they were able to reduce potential error by up to 78%. They believe that their approach has a wide range of applicability and can be very important for decision-making and resource allocation.
Discussion
I thought the goal of the researchers was a very valid one. Data entry is something that people do all the time but usually make quite a few errors. I always double and even triple-check when I perform data entry as errors have caused many problems for myself. Something like this has a large range of applicability, something that I applaud them for because it's been a while since i've read a paper I have felt that way.
I agree that error checking on data entry is very important
ReplyDelete