ABSTRACT
We describe a technique to analyse character-level errors in evaluations of text entry methods. Using an algorithm for sequence comparisons, we generate the set of optimal alignments between the presented and transcribed text. Percharacter errors, categorized as insertions, substitutions, or deletions, are obtained by analysing the alignments and applying a weighting factor. A detailed example using a real data set is given.
- MacKenzie, I. S. Mobile text entry using three keys, Proceedings of NordiCHI 2002 (to appear). Google ScholarDigital Library
- Soukoreff, R. W., and MacKenzie, I. S. Measuring errors in text entry tasks: An application of the Levenshtein string distance statistic, Extended Abstracts of CHI 2001. New York: ACM, 2001, 319-320. Google ScholarDigital Library
Index Terms
- A character-level error analysis technique for evaluating text entry methods
Recommendations
Measuring errors in text entry tasks: an application of the Levenshtein string distance statistic
CHI EA '01: CHI '01 Extended Abstracts on Human Factors in Computing SystemsWe propose a new technique based on the Levenshtein minimum string distance statistic for measuring error rates in text entry research. The technique obviates the need to artificially constrain subjects to maintain synchronization with the presented ...
Recent developments in text-entry error rate measurement
CHI EA '04: CHI '04 Extended Abstracts on Human Factors in Computing SystemsPreviously, we defined robust and easy-to-calculate error metrics for text entry research. Herein, we announce a software implementation of this error analysis technique. We build on previous work, by introducing two new metrics, and we extend error ...
Analyzing the input stream for character- level errors in unconstrained text entry evaluations
Recent improvements in text entry error rate measurement have enabled the running of text entry experiments in which subjects are free to correct errors (or not) as they transcribe a presented string. In these “unconstrained” experiments, it is no ...
Comments