You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

Daniel_Burfoot comments on Open thread, Sep. 19 - Sep. 25, 2016 - Less Wrong Discussion

2 Post author: DataPacRat 19 September 2016 06:34PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (92)

You are viewing a single comment's thread.

Comment author: Daniel_Burfoot 20 September 2016 05:32:51PM *  3 points [-]

I have a question for LWers who are non-native English speakers.

I am working on a software system for linguistically sophisticated analysis of English text. At the core of the system is a sentence parser. Unlike most other research in NLP, a central goal of my work is to develop linguistic knowledge and then build that knowledge into the parser. For example, my system knows that the verb ask connects strongly to subjectized infinitive phrases ("I asked him to take out the trash"), unlike most other verbs.

The system also has a nice parse visualization tool, which shows the grammatical structure of an input sentence. You can check it out here.

This work began as a research project and I am trying to figure out a way to commercialize it. One of my ideas is to use the system as a tool for helping students to learn English. Students could submit confusing sentences to the system and observe the parse tree, allowing them to understand the grammatical structure. They could also submit their own written sentences to the system, as a way of checking their grammar. Teachers of ESL students might also ask them to submit their class papers to the parser to check for obvious mistakes (apparently there are many people who can communicate well in spoken English but whose written English is full of mistakes).

I would also write up a series of articles about subtle points of English grammar, such as phrasal verbs, argument structure, verb tense, and so on. Students could then read the articles and experiment with using the relevant grammar patterns in the parser.

Does this sound like a plausible product that people would want to use? Are there products already on the market that do something similar? (I am aware of Grammarly, but it doesn't appear to show parse trees).

Comment author: Gunnar_Zarncke 20 September 2016 10:12:45PM 2 points [-]

Poll for it:

I'm a native speaker

Such a tool to visualize parse trees would be/have been helpful.

Submitting...

Comment author: Daniel_Burfoot 24 September 2016 03:28:30AM 1 point [-]

Thanks to Gunnar for setting up the poll and also to all who answered.

Comment author: Gunnar_Zarncke 24 September 2016 03:13:34PM *  0 points [-]

You should look at the correlations in the raw data really.

Also: Polls are easy. You should have created your own really.

See https://wiki.lesswrong.com/wiki/Comment_formatting#Polls

Comment author: MrMind 26 September 2016 09:10:38AM *  0 points [-]

How would you use the grammar visualization tool to aid study? Many people answered "unsure" to the poll because it's not clear how it should be used, or "Not really" because the first uses they thought about were not helpful.
You should give the user the guidelines on how to better consume your product.

Usually needs --> tools. Yours seems a case of inverted implication.