Edit Distance
A metric measuring the degree of difference between two strings, defined as the minimum number of insertion, deletion and substitution operations required to transform one string into another. It is commonly used in the evaluation of OCR and text recognition.
Timeline (last 90 days)
Fuzzy matching is generally implemented in engineering via edit distance, and search engines typically allow a certain edit distance threshold to find sufficiently close terms.
Relaxing the edit distance brings two side effects: more recall (possibly introducing irrelevant results) and increased computation, so the degree of fuzziness needs to be limited and combined with relevance scoring.
Edit distance (Levenshtein Distance) measures the minimum number of insertion, deletion, and substitution operations required to convert the recognition result into the standard answer
Edit distance and character accuracy are insensitive to the 'position' and 'structure' of errors, and lack a clear 'pass/fail' semantic boundary
Common methods for comparing behavioral trajectories include edit distance, sequence alignment, and embedding-based similarity computation.
Traditional structural-similarity clustering relies on edit distance or the topological similarity of control-flow graphs to merge variants, capturing only the surface form of execution paths and being unable to perceive the business meaning behind activities.
Normalized edit distance divides the raw edit distance by the length of the longer of the two strings, normalizing its range to [0, 1] and eliminating the effect of string length on the metric.
All Facts (7)
Fuzzy matching is generally implemented in engineering via edit distance, and search engines typically allow a certain edit distance threshold to find sufficiently close terms.
50%UnverifiedRelaxing the edit distance brings two side effects: more recall (possibly introducing irrelevant results) and increased computation, so the degree of fuzziness needs to be limited and combined with relevance scoring.
50%UnverifiedEdit distance (Levenshtein Distance) measures the minimum number of insertion, deletion, and substitution operations required to convert the recognition result into the standard answer
50%UnverifiedEdit distance and character accuracy are insensitive to the 'position' and 'structure' of errors, and lack a clear 'pass/fail' semantic boundary
50%UnverifiedCommon methods for comparing behavioral trajectories include edit distance, sequence alignment, and embedding-based similarity computation.
50%UnverifiedTraditional structural-similarity clustering relies on edit distance or the topological similarity of control-flow graphs to merge variants, capturing only the surface form of execution paths and being unable to perceive the business meaning behind activities.
50%UnverifiedNormalized edit distance divides the raw edit distance by the length of the longer of the two strings, normalizing its range to [0, 1] and eliminating the effect of string length on the metric.
50%