Package opennlp.spellcheck.symspell

Class SymSpellConfig

java.lang.Object

opennlp.spellcheck.symspell.SymSpellConfig

public final class SymSpellConfig
extends Object

Immutable configuration for SymSpell, created through builder().

The tunables mirror the SymSpell reference implementation:

• maxDictionaryEditDistance – the largest edit distance for which the deletes index is precomputed; queries cannot exceed this.
• prefixLength – only the first prefixLength symbols of each term are used to generate deletes, trading index size for recall on long words.
• countThreshold – the minimum corpus count for a term to be indexed.
• editDistance – the verification metric, defaulting to DamerauOSADistance.
• corpusWordCount – the corpus normalization constant N used by the Naive-Bayes word combine/split scoring in SymSpell.lookupCompound(String, int). Defaults to DERIVE_CORPUS_WORD_COUNT, which makes the engine derive N from the summed counts of the loaded dictionary so it is always corpus-correct; set it explicitly to pin N (e.g. to the full-corpus total a reference dictionary was drawn from).

See Also:

• SymSpell reference implementation (Wolf Garbe)

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	SymSpellConfig.Builder	Mutable builder for SymSpellConfig.

Field Summary

Fields

Modifier and Type	Field	Description
static final long	DERIVE_CORPUS_WORD_COUNT	Sentinel for corpusWordCount() meaning "derive N from the summed counts of the loaded dictionary" rather than pinning it to a fixed value.

Method Summary

Modifier and Type	Method	Description
static SymSpellConfig.Builder	builder()
long	corpusWordCount()
long	countThreshold()
static SymSpellConfig	defaultConfig()
EditDistance	editDistance()
int	maxDictionaryEditDistance()
int	prefixLength()

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

DERIVE_CORPUS_WORD_COUNT

public static final long DERIVE_CORPUS_WORD_COUNT

Sentinel for corpusWordCount() meaning "derive N from the summed counts of the loaded dictionary" rather than pinning it to a fixed value.

See Also:

• Constant Field Values

Method Details

maxDictionaryEditDistance

public int maxDictionaryEditDistance()

prefixLength

public int prefixLength()

countThreshold

public long countThreshold()

editDistance

public EditDistance editDistance()

corpusWordCount

public long corpusWordCount()

Returns:

the pinned corpus normalization constant N, or DERIVE_CORPUS_WORD_COUNT when N is derived from the loaded dictionary's summed counts.

builder

public static SymSpellConfig.Builder builder()

Returns:

a builder with the SymSpell reference defaults (maxDictionaryEditDistance=2, prefixLength=7, countThreshold=1, editDistance=DamerauOSADistance).

defaultConfig

public static SymSpellConfig defaultConfig()

Returns:

a configuration with all reference defaults.