get_dbpedia_uris()
has new argumenttypes
to filter results.dbpedia_spotlight_status()
without warnings if docker not available / not running #32.get_dbpedia_uris()
has new argumentsupport
#30.- The confusingly mixed usage of argument 'limit' in
dbpedia_get_wikidata_uris()
is resolved by adding a new argument 'chunksize' #29. - As a matter of consistency, argument 'limit' of
query_wikidata()
has been renamed tochunksize
#29. xml_enrich()
now adds new attributes to pre-annotated featuresget_dbpedia_uris()
method now includes argumentexpand_to_token
for subcorpus_bundles as wellmap_types_to_class()
works with the list representation in the types column- new functions
to_annotation()
,xml_enrich()
namespaced_xpath()
and methodget_dbpedia_uris()
for xml docs. - dropped argument
return_types
fromget_dbpedia_uris()
. Columntypes
is kept in output, with parsed output #24. - Error avoided when
get_dbpedia_uris()
is restricted to pre-annotated named entities and the document does not contain any (addresses issue #23). - set default
max_len
inget_dbpedia_uris()
to 5600 to avoid failing queries. get_dbpedia_uris()
optionally returns types now (addresses issue #24).- added
map_types_to_class
as an utility function to reduce types to a limited set of classes. - modified
as_subcorpus()
to makeread()
work without pre-annotated entities and in tandem withmap_types_to_class()
by avoiding hard-coding the column name "ne_type". Also added a color code to entities which are not within "PERSON", "LOCATION", "ORGANIZATION" and "MISC". - started to implement the
expand_to_token
argument which ultimately should resolve mismatches between DBpedia Spotlight's entity spans and CWB token spans (issue #26) - added
drop_inexact_annotations
argument toget_dbpedia_uris()
to control keeping or omitting inexact annotations in output data.table (see issue #26) - Error avoided when
get_dbpedia_uris()
does not detect any URI. - More telling progress messages of
wikidata_query()
anddbpedia_get_wikidata_uris()
. dbpedia_get_wikidata_uris()
implemented forcharacter
strings.- Preliminary implementation of
dbpedia_get_wikidata_uris()
for 'corpus' objects from package quanteda. - New function
add_wikidata_uris()
to add wikidata URIs to a table with DBpedia URIs. wikidata_query()
is a method now, result does not include columns "key and keyLabel" any more. The column with values queries from Wikidata is ID now (not "label", as previously).- New auxiliary function
sparql_query()
replacesSPARQL::SPARQL()
and is basis for dropping packages SPARQL, XML and RCurl as dependencies. Package xml2 is a new dependency. - Default value of argument
max_len
is now 5680. - Bugged result for
get_dbpedia_uris()
if s_attribute = NULL corrected #18.
- New function
dbpedia_spotlight_status()
.
- New auxiliary function
as_chunks()
. - New function
dbpedia_get_wikidata_uris()
. - New function
wikidata_query()
as high-level wrapper forWikidataQueryServiceR::query_wikidata()
. - Method
get_dbpedia_links()
renamed toget_dbpedia_uris()
, return value is now adata.table
, argumentmw
of methodget_dbpedia_links()
forsubcorpus
objects renamed ass_attribute
. - Auxiliary function
as_subcorpus()
to turndata.table
with DBpedia URIs intosubcorpus
that can be used for annotation.
get_dbpedia_links()
returnssubcorpus
with information in slotannotations
.
- Vignette with a very, very simple example.
- Added 'SystemRequirements: docker' to DESCRIPTION