#include <mbma_mod.h>
|
static std::map< std::string, std::string > | TAGconv |
|
static std::string | mbma_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbma-nl" |
|
static std::string | pos_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbpos-cgn" |
|
static std::string | clex_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbpos-clex" |
|
◆ Mbma()
Mbma::Mbma |
( |
TiCC::LogStream * |
errlog, |
|
|
TiCC::LogStream * |
dbglog = 0 |
|
) |
| |
|
explicit |
create an Mbma classifier object
- Parameters
-
errlog | the LogStream to use for Error messages |
dbglog | the LogStream to use for Debug messages |
◆ ~Mbma()
◆ add_morphemes()
void Mbma::add_morphemes |
( |
const std::vector< folia::Word * > & |
wv, |
|
|
const frog_data & |
fd |
|
) |
| const |
◆ add_provenance()
void Mbma::add_provenance |
( |
folia::Document & |
doc, |
|
|
folia::processor * |
main |
|
) |
| const |
add provenance information to the FoLiA document
- Parameters
-
doc | the foLiA document we are working on |
main | the main processor (presumably Frog) we want to add a new one to |
◆ assign_compounds()
void Mbma::assign_compounds |
( |
| ) |
|
add compound information to the result
◆ Classify() [1/2]
void Mbma::Classify |
( |
const icu::UnicodeString & |
word | ) |
|
◆ Classify() [2/2]
◆ clearAnalysis()
void Mbma::clearAnalysis |
( |
| ) |
|
◆ execute()
vector< Rule * > Mbma::execute |
( |
const icu::UnicodeString & |
word, |
|
|
const std::vector< std::string > & |
classes |
|
) |
| |
attempt to find matching Rules
- Parameters
-
word | a word to check |
classes | the Timbl classifications |
- Returns
- 0 or more matching Rules
◆ filterHeadTag()
void Mbma::filterHeadTag |
( |
const std::string & |
head | ) |
|
reduce the Mbms analysis by removing all solutions where the head is not matched
- Parameters
-
head | the head-tag that is required matching does not mean equality. We are a forgivingful in the sense that N matches PN
A matches B and vv
A matches V
|
◆ filterSubTags()
void Mbma::filterSubTags |
( |
const std::vector< std::string > & |
feats | ) |
|
reduce the analyses set based on sub-features
- Parameters
-
feats | a list of subfeatures when a candidate Rule has inflexion it should match a feature. |
Other criteria: only take the highest confidence, and remove Rules that yield the same analysis.
Example:
The word 'appel' is according to Mbma an N with inflection e ('enkelvoud')
The tagger will assign an N too, with features [soort,ev,basis,zijd,stan]
one of the features, 'ev' will match the 'e' inflection after translation
so this is a good reading
◆ getPrettyResults()
vector< pair< string, string > > Mbma::getPrettyResults |
( |
| ) |
const |
◆ getResult()
vector< string > Mbma::getResult |
( |
| ) |
const |
◆ getResults()
vector< pair< string, string > > Mbma::getResults |
( |
| ) |
const |
◆ init()
bool Mbma::init |
( |
const TiCC::Configuration & |
config | ) |
|
initialize the Mbma analyzer using the config
- Parameters
-
config | the Configuration to use |
- Returns
- true when no problems are detected
◆ matchRule()
Rule * Mbma::matchRule |
( |
const std::vector< std::string > & |
ana, |
|
|
const icu::UnicodeString & |
word |
|
) |
| |
attempt to match an Analysis on a word
- Parameters
-
ana | one analysis result, expanded from the Timbl classifier |
word | a Unicode Word to check |
- Returns
- a matched Rule or 0
◆ setDeepMorph()
void Mbma::setDeepMorph |
( |
bool |
b | ) |
|
|
inline |
◆ version()
std::string Mbma::version |
( |
| ) |
const |
|
inline |
◆ clex_tagset
string Mbma::clex_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbpos-clex" |
|
static |
◆ mbma_tagset
string Mbma::mbma_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbma-nl" |
|
static |
◆ pos_tagset
string Mbma::pos_tagset = "http://ilk.uvt.nl/folia/sets/frog-mbpos-cgn" |
|
static |
◆ TAGconv
map< string, string > Mbma::TAGconv |
|
static |
The documentation for this class was generated from the following files: