12.5 Decomposing Signatures Software

New data mining software has been developed for handwriting identification without the bias that often accompanies human analysis of handwriting. Who wrote the Jon-Benet Ramsey ransom note? A computer program in development at the University at Buffalo may soon be able to assist in answering such a question. It is currently 98% effective in determining the authorship of handwritten documents.

The Center of Excellence in Document Analysis and Recognition (CEDAR) project funded by the National Institute of Justice has produced the first software program designed to develop computer-assisted handwriting-analysis tools for forensic applications. In criminal cases, the question of who penned a ransom note or forged a check is now the domain of human handwriting analysts. Because they are human, even the best graphologists cannot claim complete objectivity. The University of Buffalo's software is the first that can identify who wrote a particular document based on purely scientific criteria.

"A human expert may put in his or her own bias even unconsciously," says Sargur Srihari, Ph.D., principal investigator and SUNY Distinguished Professor in the Department of Computer Science and Engineering in the College of Arts and Sciences and the School of Engineering and Applied Sciences at University of Buffalo. "We have built the foundation for a handwriting analysis system that will quantify performance and increase confidence in determining a writer's identity. This is about validating individuality in handwriting," Srihari notes. "The idea that everyone's handwriting is different is taken for granted. What we have done is to develop purely scientific criteria for that premise."

It is the first time researchers have attempted to do pattern recognition based on a large database of handwriting and using a totally automated means of measuring specific features of human handwriting, according to Srihari, who also is the director of CEDAR. CEDAR is primarily devoted to a new set of pattern-recognition technologies that can recognize and read handwriting. It was CEDAR's expertise in developing systems that can read and interpret handwritten addresses on envelopes for the U.S. Postal Service—now used around the world—that attracted interest from the National Institute of Justice, which funded the work with a $428,000 grant.

Providing a scientific basis for establishing the individuality of handwriting has become essential for admitting handwriting evidence in U.S. courts due to a number of recent rulings concerning expert testimony, Srihari says; "In this project, we are developing a technology whose job it is to authenticate documents." The University of Buffalo researchers developed the software by first collecting a database of more than 1,000 samples of handwriting from a pool of individuals representing a microcosm of the U.S. population in terms of gender, age, and ethnicity.

Multiple samples of handwriting were taken from subjects, each of whom was asked to write the same series of documents in cursive. Instead of analyzing the documents visually, the way a human expert would, Srihari explains, the researchers deconstructed each sample, extracting features from the writing, such as measuring the shapes of individual characters, descenders, and the spaces between lines and words. The researchers then ran the samples through their software program.

"We tested the program by asking it to determine which of two authors wrote a particular sample, based on measurable features," says Srihari. "The program responded correctly 98 percent of the time." Human experts look for arcades and garlands, features that may distinguish one person's penmanship from another's, he explains. The current software should be able to conduct that type of advanced analysis within the year.

The goal of authenticating documents in criminal cases is usually to determine whether or not a particular suspect wrote the document in question. The scientific approach that Srihari and his colleagues are developing, however, also may be useful in establishing individuality, as with DNA, fingerprints, or facial features, in the emerging field of biometrics, which is the automated identification of a person based on precise measurements of physiological or behavioral characteristics. For more information on CEDAR, go to its Web site at http://www.cedar.buffalo.edu/index.html.

The following case study is again provided in its original version. The author would like to thank Lars Kangas for his valuable assistance in providing the materials relating to his work with the CATCH investigative data mining project.