Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Is this dependnecy label correct ?

See original GitHub issue

I’ve tried a sentence “僕がリンゴを買った。” and got curious result.

% echo '僕がリンゴを買った。' | ginza
mode is C
disabling sentence separator
# text = 僕がリンゴを買った。
1       僕      僕      PRON    代名詞  _       5       iobj    _       BunsetuBILabel=B|BunsetuPositionType=SEM_HEAD|SpaceAfter=No|NP_B
2       が      が      ADP     助詞-格助詞     _       1       case    _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
3       リンゴ  林檎    NOUN    名詞-普通名詞-一般      _       5       obj     _       BunsetuBILabel=B|BunsetuPositionType=SEM_HEAD|SpaceAfter=No|NP_B
4       を      を      ADP     助詞-格助詞     _       3       case    _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
5       買っ    買う    VERB    動詞-一般       _       0       root    _       BunsetuBILabel=B|BunsetuPositionType=ROOT|SpaceAfter=No
6       た      た      AUX     助動詞  _       5       aux     _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
7       。      。      PUNCT   補助記号-句点   _       5       punct   _       BunsetuBILabel=I|BunsetuPositionType=CONT|SpaceAfter=No

The correct dep label of “僕” is nsubj, isn’t it ? The result of “太郎がリンゴを買った。” seems right.

% echo '太郎がリンゴを買った。' | ginza
mode is C
disabling sentence separator
# text = 太郎がリンゴを買った。
1       太郎    太郎    PROPN   名詞-固有名詞-人名-名   _       5       nsubj   _       BunsetuBILabel=B|BunsetuPositionType=SEM_HEAD|SpaceAfter=No|NP_B|NE=PERSON_B
2       が      が      ADP     助詞-格助詞     _       1       case    _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
3       リンゴ  林檎    NOUN    名詞-普通名詞-一般      _       5       obj     _       BunsetuBILabel=B|BunsetuPositionType=SEM_HEAD|SpaceAfter=No|NP_B
4       を      を      ADP     助詞-格助詞     _       3       case    _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
5       買っ    買う    VERB    動詞-一般       _       0       root    _       BunsetuBILabel=B|BunsetuPositionType=ROOT|SpaceAfter=No
6       た      た      AUX     助動詞  _       5       aux     _       BunsetuBILabel=I|BunsetuPositionType=SYN_HEAD|SpaceAfter=No
7       。      。      PUNCT   補助記号-句点   _       5       punct   _       BunsetuBILabel=I|BunsetuPositionType=CONT|SpaceAfter=No

Issue Analytics

State:
Created 4 years ago
Comments:5 (4 by maintainers)

Top GitHub Comments

1reaction

hiroshi-matsuda-ritcommented, Jan 18, 2020

I just researched all the occasions of 買った and those subjects in the training data-set of UD_Japanese-BCCWJ. All most all of those sentences had no subject (zero anaphor) and few sentences had non-zero subject with isubj dependency label, but I found a sentence taking a subject with nsubj label.

I think that 買った is a good example to evaluated the model robustness on zero anaphor context. I’d like to keep this issue opened for further research.

1reaction

hiroshi-matsuda-ritcommented, Nov 10, 2019

@h-asano Thank you for reporting parsing error example. I think the dependency label for those contexts should be nsubj. Actually, the latest version of the UD_Japanese-BCCWJ may contain some incoherent dep label for ‘僕が’ type context. I’d like to report this issue to Japanese UD team.

Top Results From Across the Web

Dependency Parser or how to find syntactic neighbours of a ...

An accurate Dependency Parser recognizes the dependencies and relations between words well. Two Metrics (scores) are useful for this: - ...

Part-of-Speech(POS) Tag | Dependency Parsing

This tag is assigned to the word which acts as the head of many words in a sentence but is not a child...

Dependency Parsing - Stanford University

Labeled attachment refers to the proper assignment of a word to its head along with the correct dependency relation. Unlabeled attachment simply looks...

Dependency parsing - NLP-progress

Dependency parsing is the task of extracting a dependency parse of a sentence that ... while LAS requires a semantic correct label for...

Making Dependency Labeling Simple, Fast and Accurate

This work addresses the task of dependency labeling—assigning labels to an (unlabeled) dependency tree. We employ and extend a.