COACH - Cumulative Online Algorithm for Classification of Handwriting Deficiencies

In this paper we present COACH - a Cumulative Online Algorithm for Classification of Handwriting deficiencies. A description of our algorithm along with a performance evaluation of COACH on real data is provided. COACH is an innovative algorithm designed for building an online handwriting evaluation tool to be used for classifying and remediating handwriting deficiencies. COACH adapts learning and data mining techniques from AI to handwriting deficiency classification in an innovative fashion. Until now handwriting classification has been performed manually by trained therapists causing expensive and subjective evaluation. This application lowers the cost of evaluation, increases objectiveness, and enables repeated testing that can accompany therapy. COACH is evaluated on real data obtained from children with poor handwriting using a digitizer tablet. Results show that COACH manages to successfully differentiate between poor to proficient handwriting. Differentiation is obtained even after using data from only a few words. These results prove that COACH is a promising emerging application for online evaluation.

[1]  Tao Qin,et al.  Fast Large-Scale Spectral Clustering by Sequential Shrinkage Optimization , 2007, ECIR.

[2]  Hongxing He,et al.  Feature Selection for Temporal Health Records , 2001, PAKDD.

[3]  Sargur N. Srihari,et al.  Offline Signature Verification And Identification Using Distance Statistics , 2004, Int. J. Pattern Recognit. Artif. Intell..

[4]  Ichigaku Takigawa,et al.  A Probabilistic Model for Clustering Text Documents with Multiple Fields , 2007, ECIR.

[5]  Claude Sammut,et al.  Constructive Induction for Classifying Time Series , 2004, ECML.

[6]  Jing Liu,et al.  Multivariate time series prediction via temporal classification , 2002, Proceedings 18th International Conference on Data Engineering.

[7]  Sara Rosenblum,et al.  Computerized temporal handwriting characteristics of proficient and non-proficient handwriters. , 2003, The American journal of occupational therapy : official publication of the American Occupational Therapy Association.

[8]  Nicole Vincent,et al.  Handwriting and signature: one or two personality identifiers? , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[9]  Huan Liu,et al.  Feature Selection for Classification , 1997, Intell. Data Anal..

[10]  Ian Witten,et al.  Data Mining , 2000 .

[11]  Claus Bahlmann,et al.  Directional features in online handwriting recognition , 2006, Pattern Recognit..

[12]  Francesco Carlo Morabito,et al.  Fuzzy neural identification and forecasting techniques to process experimental urban air pollution data , 2003, Neural Networks.

[13]  Sara Rosenblum,et al.  The in Air Phenomenon: Temporal and Spatial Correlates of the Handwriting Process , 2003, Perceptual and motor skills.

[14]  Mohammed J. Zaki,et al.  PlanMine: Predicting Plan Failures Using Sequence Mining , 1998, Artificial Intelligence Review.

[15]  Sriganesh Madhvanath,et al.  An approach to identify unique styles in online handwriting recognition , 2005, Eighth International Conference on Document Analysis and Recognition (ICDAR'05).

[16]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .