 |
ML Newsnew | past | best | rss | submit | about |
| A logical word embedding for learning grammar (arxiv.org) | |
| 2 points by PaulHoule on Nov 5, 2023 | hide | past | pdf | discuss on HN | |
|
In plain words: A model that learns word categories and grammar rules from raw text with no labels, instead of relying on hand-written rules, and prints its findings in plain words. It works from as few as 100 sentences and builds new sentences in steps you can see. AbstractWe introduce the logical grammar emdebbing (LGE), a model inspired by pregroup grammars and categorial grammars to enable unsupervised inference of lexical categories and syntactic rules from a corpus of text. LGE produces comprehensible output summarizing its inferences, has a completely transparent process for producing novel sentences, and can learn from as few as a hundred sentences. | |
| add comment on HN |