HNHacker News
TopNewBestAskShowJobs

rspeer

5,317 karma · joined January 4, 2010

submissionscomments

How to make a racist AI without really trying

blog.conceptnet.io·6 pts·rspeer·
0

ConceptNet Numberbatch 17.04: better, less-stereotyped word vectors

blog.conceptnet.io·6 pts·rspeer·
0

How Luminoso made ConceptNet into the best word vectors, and won at SemEval

blog.conceptnet.io·3 pts·rspeer·
0

Show HN: Announcing ConceptNet 5.5 and conceptnet.io

blog.conceptnet.io·5 pts·rspeer·
0

Wordfreq 1.5: More data, more languages, more accuracy

blog.conceptnet.io·2 pts·rspeer·
0

Yes, people do want pre-computed word embeddings

blog.conceptnet.io·1 pts·rspeer·
0

Cramming for the test set: We need better ways to evaluate analogies

blog.conceptnet.io·3 pts·rspeer·
0

Conceptnet Numberbatch: A new name for the best word embeddings you can download

blog.conceptnet.io·2 pts·rspeer·
0

An Introduction to the ConceptNet Vector Ensemble

blog.luminoso.com·16 pts·rspeer·
0

Wordfreq: Open source and open data about word frequencies

blog.luminoso.com·2 pts·rspeer·
0

Ftfy (fixes text for you) 4.0: changing less and fixing more

blog.luminoso.com·2 pts·rspeer·
0

Natural language can be such an ass headache

blog.luminoso.com·1 pts·rspeer·
0

Show HN: Langcodes – A Python library for working with language codes

github.com·5 pts·rspeer·
0

Python-ftfy: Given Unicode text, make its representation possibly less broken

github.com·101 pts·rspeer·
40

U+DEADBEEF: Why you shouldn't trust arbitrary text encodings

rspeer.github.io·1 pts·rspeer·
0

Picking a March Madness bracket using natural-language text

blog.luminoso.com·1 pts·rspeer·
0

Emoji are more common than hyphens

blog.luminoso.com·2 pts·rspeer·
0

Ftfy 3.0 — makes Unicode text less broken

blog.luminoso.com·2 pts·rspeer·
0

Show HN: Fixing common Unicode mistakes with Python — after they've been made

blog.lumino.so·4 pts·rspeer·
0