Work Packages

• Theory for weighted Levenshtein automata
• Ordering correction candidates according their word frequencies
• Representative corpora collection
• Word frequencies analysis
• Analysis of recognition error risk
• Construction of Bulgarian, Russian, German and English OCR dictionaries
• Construction of Bulgarian-Russian-German-English consolidated OCR Dictionary
• Collection of font samples of representative documents
• Analysis of the font samples for symbol dependent recognition errors
• Statistical analysis of representative corpora for word collocation