Statistics

This was made after analyzing ~50 Swedish riksdagen open data documents out of 600k in total.

Swedish

Lexical categories

540k links in total

Number of sentences

Number of sentences with a langdetect score > 0.7

Number of sentences with at least one linked entity

Note: only one document was analyzed since the PR was merged

Rawtokens with a langdetect score > 0.7 (we accept > 0.4)

Garbage tokens

count_entities_in_sv_sentences_group_by_label.sql

Average entity count per document

Note: only one document was analyzed since the PR was merged

Entity count per document

Note: only one document was analyzed since the PR was merged

Database size

Source

Queries are found in th sql subfolder, see https://github.com/dpriskorn/riksdagen_sentences/tree/add_lookup_endpoint/sql