Robustness Analysis of Adaptive Chinese Input Methods

Mike Tian-Jian Jiang, Cheng-Wei Lee, Chad Liu, Yung-Chun Chang, Wen Lian Hsu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This work proposes a novel metric, Maximally Amortized Cost (MAC), for cost evaluations of error correction of predictive Chinese input methods (IMs). With a series of real-time simulation, user correction behaviors are analyzed by estimating generalized backward compatibility of adaptive Chinese IMs. Comparisons between three IMs by using MAC with different context lengths report empirical factors of context length for improving predictive IMs. The error-tolerance level—Futile Effort, Beneficial Effort and Utility—of adaptive IMs is also proposed and analyzed.
Original languageEnglish
Title of host publicationWorkshop on Advances in Text Input Methods (WTIM)
Publication statusPublished - Dec 2011

Fingerprint

Costs
Error correction

Cite this

Jiang, M. T-J., Lee, C-W., Liu, C., Chang, Y-C., & Hsu, W. L. (2011). Robustness Analysis of Adaptive Chinese Input Methods. In Workshop on Advances in Text Input Methods (WTIM)

Robustness Analysis of Adaptive Chinese Input Methods. / Jiang, Mike Tian-Jian; Lee, Cheng-Wei ; Liu, Chad; Chang, Yung-Chun; Hsu, Wen Lian.

Workshop on Advances in Text Input Methods (WTIM). 2011.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jiang, MT-J, Lee, C-W, Liu, C, Chang, Y-C & Hsu, WL 2011, Robustness Analysis of Adaptive Chinese Input Methods. in Workshop on Advances in Text Input Methods (WTIM).
Jiang MT-J, Lee C-W, Liu C, Chang Y-C, Hsu WL. Robustness Analysis of Adaptive Chinese Input Methods. In Workshop on Advances in Text Input Methods (WTIM). 2011
Jiang, Mike Tian-Jian ; Lee, Cheng-Wei ; Liu, Chad ; Chang, Yung-Chun ; Hsu, Wen Lian. / Robustness Analysis of Adaptive Chinese Input Methods. Workshop on Advances in Text Input Methods (WTIM). 2011.
@inproceedings{4431bcd9f52e4e3fb52f35f40022f3fd,
title = "Robustness Analysis of Adaptive Chinese Input Methods",
abstract = "This work proposes a novel metric, Maximally Amortized Cost (MAC), for cost evaluations of error correction of predictive Chinese input methods (IMs). With a series of real-time simulation, user correction behaviors are analyzed by estimating generalized backward compatibility of adaptive Chinese IMs. Comparisons between three IMs by using MAC with different context lengths report empirical factors of context length for improving predictive IMs. The error-tolerance level—Futile Effort, Beneficial Effort and Utility—of adaptive IMs is also proposed and analyzed.",
author = "Jiang, {Mike Tian-Jian} and Cheng-Wei Lee and Chad Liu and Yung-Chun Chang and Hsu, {Wen Lian}",
year = "2011",
month = "12",
language = "English",
booktitle = "Workshop on Advances in Text Input Methods (WTIM)",

}

TY - GEN

T1 - Robustness Analysis of Adaptive Chinese Input Methods

AU - Jiang, Mike Tian-Jian

AU - Lee, Cheng-Wei

AU - Liu, Chad

AU - Chang, Yung-Chun

AU - Hsu, Wen Lian

PY - 2011/12

Y1 - 2011/12

N2 - This work proposes a novel metric, Maximally Amortized Cost (MAC), for cost evaluations of error correction of predictive Chinese input methods (IMs). With a series of real-time simulation, user correction behaviors are analyzed by estimating generalized backward compatibility of adaptive Chinese IMs. Comparisons between three IMs by using MAC with different context lengths report empirical factors of context length for improving predictive IMs. The error-tolerance level—Futile Effort, Beneficial Effort and Utility—of adaptive IMs is also proposed and analyzed.

AB - This work proposes a novel metric, Maximally Amortized Cost (MAC), for cost evaluations of error correction of predictive Chinese input methods (IMs). With a series of real-time simulation, user correction behaviors are analyzed by estimating generalized backward compatibility of adaptive Chinese IMs. Comparisons between three IMs by using MAC with different context lengths report empirical factors of context length for improving predictive IMs. The error-tolerance level—Futile Effort, Beneficial Effort and Utility—of adaptive IMs is also proposed and analyzed.

UR - https://www.researchgate.net/publication/229086622_Robustness_Analysis_of_Adaptive_Chinese_Input_Methods

M3 - Conference contribution

BT - Workshop on Advances in Text Input Methods (WTIM)

ER -