Patent ReferencesRegional context maximum likelihood error correction for OCR, keyboard, and the like Method and apparatus for context-aided recognition Graphic word spelling correction using automated dictionary comparisons with phonetic skeletons System to achieve automatic recognition of linguistic strings Means for resolving ambiguities in text based upon character context Neural network model in pattern recognition using probabilistic contextual information Means for resolving ambiguities in text based upon character context Pattern recognition system having inter-pattern spacing correction Knowledge-based character recognition Method for correcting spelling using error frequencies InventorsApplicationNo. 933183 filed on 09/18/1997US Classes:715/531, Text382/309, EDITING, ERROR CHECKING, OR CORRECTION (E.G., POSTRECOGNITION PROCESSING)382/310, Correcting alphanumeric recognition errors382/321, Optical (e.g., OCR)715/530Edit, composition, or storage controlExaminersPrimary: Hong, Stephen S.Assistant: Stevens, Roberta Attorney, Agent or FirmInternational ClassG06F 017/27AbstractA computer program which causes a computer to correct a uniform resource identifier (URI) in a noisy source document. The program finds and corrects potential errors within a URI before turning the URI into a hyperlink. Testing the corrected URI is done by seeking the resource described by the corrected URI. Testing the URI also includes parsing the URI, identifying potential syntax errors within each portion of the URI, creating alternative URI combinations, and prioritizing the alternative URI combinations. Syntax errors corrected include incorrect protocol, incorrect or missing component separator characters, incorrect spacing, incorrect or missing dot character, and alphanumeric character replacement. | |