Gradually disappearing pinyin - web app

[webmagnets]

I've developed Slowmersion.com - It is a web app for reading Chinese with pinyin and then gradually weaning you off of the pinyin based on the words you know well.

 

It is in beta and I am interested in feedback on the web app, the videos explaining it, and whatever else.

[imron]

Looks interesting.  How are you doing your segmentation and pinyin generation, and how do you account for different readings of a single character/word.

 

Also, I would de-emphasise the "one man team" part of your landing page.

[webmagnets]

I'm using https://www.npmjs.com/package/nodejieba for segmentation.

Pinyin comes from CC-CEDICT

If a word has multiple pronunciations, then the pinyin will be purple ("purple" rhymes-ish with "multiple").

 

Thanks for the input on the "one man team" stuff. I've been saying "we" for so long when it is really only "I". I was hoping the world could handle some honesty in that regard. You are probably right though.

[Angelina]

I was reading something on how native speakers of Chinese would segment longer items, while speakers of Chinese as a second language would make shorter items when reading the same text. This is only one study though, don't take it for granted. 

Have you considered this? At least the possibility of different segmentation habits between learners and native speakers. 

[webmagnets]

2 minutes ago, Angelina said:

Have you considered this?

 

I hadn't considered that.

 

It seems that nodejieba is geared toward native speakers. I had developed my own segmentation algorithm, but it was way too slow. I decided focus on my core features instead of trying to reinvent the wheel.

[Angelina]

NLP 

Linguistics doesn't sell

加油!

[imron]

2 hours ago, webmagnets said:

Thanks for the input on the "one man team" stuff. I've been saying "we" for so long when it is really only "I". I was hoping the world could handle some honesty in that regard.

Honestly, for products or services, I think 'we' is acceptable even if it's just run by one person.

 

Maybe though you can go for something more ambiguous such as

 

Slowmersion was created and developed by 'webmagnets'.

 

That will still work well if the site takes off and you ever need to take on staff.

[tiantian]

I really like it! I like how it automatically keeps track of the words you did not look up while reading. The segmentation is also much better than most other reading apps. I also like how it is web based and works pefectly on my ipad. Really great work! The "Show word in sentence where you last saw it" feature, which I think is a great idea, does not seem to work though. Is it implemented yet?

 

What I would like to be added in the future is perhaps a list/count/statistic of the words you have already learned.

 

Regards,

 

Tiantian

[tiantian]

Oh and it would also be nice to be able to kind of pre-select a known words list, perhaps HSK-based. This way intermediate/advanced learners don't have to start from scratch when pasting their texts. In my case I would just select HSK 1-4 and go from there. 

[webmagnets]

Thank you very much @tiantian for your feedback, feature suggestions and the heads up about the "sentence immersion" feature not working. I think I may have got it working now.

 

I want to offer charts/statistics about word learning progress in the future.

 

Also, I have been thinking about how to quickly get the app up to speed on the user's reading level. Your idea might work.

[mlescano]

4 hours ago, webmagnets said:

I have been thinking about how to quickly get the app up to speed on the user's reading level.

 

Maybe allow one to import a wordlist? I use Pleco flashcards, and I've sometimes exported wordlists for filtering Chinese Word Extractor results. Other people might use Anki.

[atongmu]

I don't like the religious nature of this program. No sowing of seeds please.

[roddy]

Eh? What religious nature?

[Flickserve]

10 hours ago, mlescano said:

Maybe allow one to import a wordlist? I use Pleco flashcards, and I've sometimes exported wordlists for filtering Chinese Word Extractor results. Other people might use Anki.

 

On the other hand, the first setup can work as a real test of one's pinyin....

[webmagnets]

4 hours ago, Flickserve said:

On the other hand, the first setup can work as a real test of one's pinyin....

 

What do you mean?

[Flickserve]

1 hour ago, webmagnets said:

What do you mean?

 

People may sometimes overestimate their knowledge or not realised they have memory decay.

 

When you load a wordlist, you can test them on this list. If they write any wrong pinyin, your program keeps showing the pinyin.

 

 

[atongmu]

@roddy the links are all religious in nature.

 

It seems he has removed some of them, but this one is still up.

 

actually the program is quite interesting, but I don't fundamentalists spreading their beliefs in this way.

 

 

[webmagnets]

I don't control what my users make public.

[atongmu]

Ok but that may be something to think about.

 

I have put an article from Chairmans Bao to try out. So I will give it a go and give you some more feedback. 

[imron]

In general I would say it's not a good idea to control too much the content that users generate (illegal material excepting), but perhaps allowing people to filter/ignore posts by certain users would solve that problem.

 

People can still post what they want, and people who don't like it can filter it away.