Lexical postcorrection of OCR-results:the web as a dynamic secondary dictionary?

Strohmaier, Christian; Ringlstetter, Christoph; Schulz, Klaus U.; Mihov, Stoyan

doi:10.1109/icdar.2003.1227833

Cited by 20 publications

(26 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this experiment, optimal results were obtained using one single "dynamic" subdictionary which was obtained via automated analysis of the vocabulary of web pages associated with the field of botany. Details of the general method for dynamic dictionary construction are described in [15]. For defining an optimal score, small values of α turned out to be useful.…”

Section: Evaluation Resultsmentioning

confidence: 99%

“…We also use dynamic dictionaries obtained via analysis of web vocabulary, cf. [15]. The standard Levenshtein distance 1 is used as basic distance measure.…”

Section: Basic Resourcesmentioning

confidence: 99%

“…For example, in earlier experiments with German documents we only reached a lexical coverage between 93.07% and 98.42% even with very large dictionaries that are enriched with vocabulary from thematic web pages ( [15]). Hence, for automated correction tasks a correction candidate should only be used to replace the original OCR result in the presence of additional confidence.…”

Section: Introductionmentioning

confidence: 99%

“…Examples are dictionaries for proper names, geographic names, for terminological expressions from fixed thematic areas, acronyms, etc. Another suggestion, discussed in [15], is to derive for each application an additional "dynamic" dictionary that is obtained via analysis of web pages from the given thematic area retrieved by web search engines. In a given application we may then look for an optimal combination of a base dictionary with special dictionaries and a dynamic dictionary.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A visual and interactive tool for optimizing lexical postcorrection of OCR results

Strohmaier

Ringlstetter²,

Schulz³

et al. 2003

2003 Conference on Computer Vision and Pattern Recognition Workshop

View full text Add to dashboard Cite

show abstract

Section: Evaluation Resultsmentioning

confidence: 99%

“…We also use dynamic dictionaries obtained via analysis of web vocabulary, cf. [15]. The standard Levenshtein distance 1 is used as basic distance measure.…”

Section: Basic Resourcesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A visual and interactive tool for optimizing lexical postcorrection of OCR results

Strohmaier

Ringlstetter²,

Schulz³

et al. 2003

2003 Conference on Computer Vision and Pattern Recognition Workshop

View full text Add to dashboard Cite

show abstract

“…In Section 3, we discuss our formulation of the scene text recognition problem as an integer program, including the criterion of optimization, and using equivalence and differences of character pairs as constraints in the integer program. In Section 4, we present a method of using a search engine to correct some errors in the signs in our database, similar to the work of others [20,10]. In Section 5 we detail our experiments, and present results.…”

Section: Prior Work In Similaritymentioning

confidence: 99%

Enforcing similarity constraints with integer programming for better scene text recognition

2011

View full text Add to dashboard Cite

The recognition of text in everyday scenes is made difficult by viewing conditions, unusual fonts, and lack of linguistic context. Most methods integrate a priori appearance information and some sort of hard or soft constraint on the allowable strings. Weinman and showed that the similarity among characters, as a supplement to the appearance of the characters with respect to a model, could be used to improve scene text recognition. In this work, we make further improvements to scene text recognition by taking a novel approach to the incorporation of similarity. In particular, we train a similarity expert that learns to classify each pair of characters as equivalent or not. After removing logical inconsistencies in an equivalence graph, we formulate the search for the maximum likelihood interpretation of a sign as an integer program. We incorporate the equivalence information as constraints in the integer program and build an optimization criterion out of appearance features and character bigrams. Finally, we take the optimal solution from the integer program, and compare all "nearby" solutions using a probability model for strings derived from search engine queries. We demonstrate word error reductions of more than 30% relative to previous methods on the same data set.

show abstract

Character confusion versus focus word-based correction of spelling and OCR variants in corpora

Reynaert

2010

IJDAR

View full text Add to dashboard Cite

We present a new approach based on anagram hashing to handle globally the lexical variation in large and noisy text collections. Lexical variation addressed by spelling correction systems is primarily typographical variation. This is typically handled in a local fashion: given one particular text string some system of retrieving near-neighbors is applied, where near-neighbors are other text strings that differ from the particular string by a given number of characters. The difference in characters between the original string and one of its retrieved near-neighbors constitutes a particular character confusion. We present a global way of performing this action: for all possible particular character confusions given a particular edit distance, we sequentially identify all the pairs of text strings in the text collection that display a particular confusion. We work on large digitized corpora, which contain lexical variation due to both the OCR process and typographical or typesetting error and show that all these types of variation can be handled equally well in the framework we present. The character confusion-based prototype of Text-Induced Corpus Clean-up (ticcl) is compared to its focus word-based counterpart and evaluated on 6 years' worth of digitized Dutch Parliamentary documents. The character confusion approach is shown to gain an order of magnitude in speed on its word-based counterpart on large corpora. Insights gained about the useful contribution of global corpus variation statistics are shown to also benefit the more traditional word-based approach to spelling correction. Final tests on a held-out set comprising the 1918 edition of the Dutch daily newspaper 'Het Volk' show that the system is not sensitive to domain variation.

show abstract

Lexical postcorrection of OCR-results:the web as a dynamic secondary dictionary?

Abstract: Postcorrection of OCR-results for text documents is usu-

Cited by 20 publications

References 10 publications

A visual and interactive tool for optimizing lexical postcorrection of OCR results

A visual and interactive tool for optimizing lexical postcorrection of OCR results

Enforcing similarity constraints with integer programming for better scene text recognition

Character confusion versus focus word-based correction of spelling and OCR variants in corpora

Contact Info

Product

Resources

About