It counts the differences between two strings, where we would
count a difference not only when strings have different characters but when one has
a character and the other does not. It is the smallest number of insertions, deletions,
and substitutions required to change one string or tree into another. An H(m x n)
algorithm computes the distance between strings, where m and n are the lengths of
the strings (http://www.nist.gov/dads/HTML/levenshtein.html).
The simple correlation matrix technique is a correlation technique. Each misspelled
word is represented by an n-dimensional feature where Hamming distance of strongly
correlated matches the most probably correct word (Cherkassky, Vassilas, Brodt,
Wagner, & Fischer, 1974).
The singular value decomposition (SVD): Correlation Matrix Technique to apply
matrix transformation techniques to simple correlation matrices in an effort to improve
spelling correction accuracy (Deerwester, Dumais, Furnas, Landauer, & Harshman,
1990). The goal of SVD is to find the most relevant similarity in lexical space.
The correction of words rests on the basis of three common phenomena: nonword
error detection, isolated word error correction, and context-dependent word correction
(Kukich, 1992).
Pages:
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791