Noun chunks not present in .noun_chunks when nouns are properly tagged.

[theSage21]

This table quickly describes the problem.

Spacy Version	pos_	tag_	dep_	noun_chunks
1.8.2	NOUN VERB ADP PROPN ADP PROPN PUNCT	WP VBD IN NNP IN NNP .	nsubj ROOT prep pobj prep pobj punct	[What, F-22, Syria]
2.0.2	NOUN VERB ADP PROPN ADP PROPN PUNCT	WP VBD IN NNP IN NNP .	nsubj ROOT prep punct punct pobj punct	[What, Syria]
2.0.5	NOUN VERB ADP PROPN ADP PROPN PUNCT	WP VBD IN NNP IN NNP .	nsubj ROOT prep punct punct pobj punct	[What, Syria]

In the old spacy an adversarial sentence "What happened for F-22 in Syria?" has .noun_chunks attribute as [what, F-22, Syria] and the new one misses F-22. The pos and tags are the same in both versions. however the dep_ is different. Would this difference reliably explain the difference in noun_chunks?

Old environment

• Python version: 2.7.12
• Platform: Linux-4.4.0-1041-aws-x86_64-with-Ubuntu-16.04-xenial
• spaCy version: 1.8.2
• Installed models: en

New environment

• Platform: Linux-4.10.0-42-generic-x86_64-with-Ubuntu-17.04-zesty
• Models: en_core_web_md, en
• Python version: 3.5.3
• spaCy version: 2.0.2

[ines]

Yes, the noun chunks depend on the part-of-speech tags and the dependency parse, so this issue likely comes down to the difference in predictions made by the parser.

I'm merging this with #3052. We've now added a master thread for incorrect predictions and related reports – see the issue for more details.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.