Following language through the stages of learning

(updated 2016-09-08)

Learning a new word generally seems to follow a typical progression. First you notice a new word, then after you've encountered it a few times, it starts to feel familiar and you decide that maybe it's worth learning. You then more actively try and commit it to memory and perhaps try using it out in the wild a few times, and eventually it joins all the other words that you feel you've mastered.

Here's what this process looks like:

Learn stages

At WordSwing, we make this process explicit. You can place any character or word into one of these three learn stages. These learn stages have the following interpretations:

  1. Discovered - This is a word that has entered your radar. You've seen or heard it a few times and it seems worth keeping your eye on.
  2. Learning - You've continued to encounter this word, and it seems like something you'd really like to know and something you would like to actively practice.
  3. Mastered - You feel confident that you know this word well.

These learn stage ratings have two main purposes:

  1. WordSwing uses these to personalize your language practice. Specifically, WordSwing will include words in the learning stage much more frequently in the more drill-like activities, and WordSwing will assume that words in the mastered stage will be helpful context with which to learn other words.
  2. You can use these to keep track of what you've learned, what you're learning, and where the gaps are, for example by using Character knowledge map.

While practicing on WordSwing, you'll have opportunities to tune what learn stage each word occupies, for example, by using the brain tool in the drill-down pane:

Learn stage drill-down tool

The three buttons at the bottom of the above screenshot let you control whether the ratings are shown to you while you study. This can be helpful for getting a sense of how much of some text you understand or for identifying words you would like to provide ratings for but haven't yet.

You can also see on the Dashboard how many words you've rated a particular learn stage and what fraction of the typical text these comprise:

Learn stage summary

So, come over to WordSwing, and start tracking your knowledge of Chinese and reaping the benefits of personalized practice!

Hope to see you soon!

Modeling capitulation - simpler is better

We have been experimenting with some of the latest and greatest in machine learning algorithms so that your practice material is most appropriate for your knowledge and ability.

Unfortunately, many of these efforts have fallen flat on their face. Our new personalization scheme is a bit more patient.

We're still a small operation and thus the data we have collected has been a bit too limited for some of these algorithms, such as neural nets, to perform well. In response, we have revamped how study material is proposed, settling on a solution that has much better performance for new students, for whom we have collected little data.

Our new personalization scheme is a bit more patient, waiting until we have a better grasp of your knowledge before heavily personalizing the material to the map we build of your knowledge. This should reduce the rather erratic behavior that sometimes was observed with our old personalization system.

At present three activities use the new personalization scheme:

  1. Character Combination
  2. Character Similarity
  3. Recall Pronunciation

Tuning to the difficulty you want

In order to compensate for less initial automatic tailoring, we have implemented a new dial which allows you to make the practice content easier or harder, depending on your current study mood:

Difficulty dial screehsnot

The return of Recall Pronunciation

If you've been around WordSwing for a while, you'll remember a very simple activity, Recall Pronunciation, that mysteriously vanished when we redesigned WordSwing to be mobile friendly. Well, it's back!

Recall pronunciation screehsnot

It's a simple way to drill your memory for pronunciation. It features auto-suggest for pinyin entry and accepts spoken pronunciation including tone sandhi changes as well as standard dictionary pronunciation.

Thanks for you patience! And if you haven't checked out WordSwing in a while, please do! And don't forget to let me know what you think.


The redesign is live

The redesign of WordSwing's website is now live.

Mobile screehsnot

What's different?

  1. It's mobile. Finding enough time to practice a language is always a challenge. Finding enough time in front of a desktop or laptop is even harder. Now that WordSwing works on mobile devices, you can take advantage of free moments when you're out and about, standing in line, waiting for a plane, riding the bus, etc.

  2. It's faster. Architectural changes to WordSwing enabled us to make the site much faster to download. This should make the load time fast enough for mobile devices. For those interested in technical details, we now serve the main application bundle using a content-delivery network that caches the files at network edge locations near you, and serve compressed files when possible.

  3. There's a whole new drill down interface. At the crux of WordSwing is the ability to drill down to learn more about language you encounter while practicing. Part of this redesign was to completely revamp the drill down interface. The old implementation required tedious mouse-hovering acrobatics to select which language component you wanted to focus on. Now we have an easy touch/drag based highlighting mechanism to quickly get to characters, words, and phrases of interest.

Why'd it take so long?

There are two answers, depending on what is meant by the question.

One answer relates to why didn't we start sooner? We whole-hartedly agree, mea culpa. Four years ago Olle was already exposing the tremendous value of learning Chinese via smartphone. Unfortunately, in an attempt to rapidly prototype and build out WordSwing's language practice functionality we made some design decisions that were at odds with touch-based access. In this rewrite, we've tried to rethink everything from a mobile-first perspective and do it right.

The other answer relates to why did the rewrite take so long? This rewrite has been occupying much of our development efforts for the past 6 months. Six months ago, when I was evaluating whether to do a rewrite I came across a blog post by Joel Spolsky (creater of StackOverflow and Trello), appropriately titled Things You should never do, part I, which discusses why a rewrite is a terrible idea. So I set out not to do a rewrite, but as I broke more and more of of my development copy of WordSwing trying to rearchitect, I got sucked in and there was no going back. What you'll now find at WordSwing is a redesign, that while not a full rewrite, still involved touching nearly all of the 80,000 lines of code that underpin WordSwing.

So, sorry this took so long! I've learned my lesson, and will try to never do it again.

What can we expect going forward?

Most importantly it means back to forward progress, and not only on software development, but also growing WordSwing's language learning content, and improving the machine-learning and modeling behind WordSwing.

From now on it should be possible to make incremental improvements, even major ones, without halting progress on all other fronts. Part of this is that the main client-side web framework I use, Ember.js, has evolved to have a very mature development cycle, complete with a set of staged release channels, semantic versioning, stability guarantees, a well-honored depracation policy, and backed by a great community of developers. This combined with architectural changes I made during this rewrite means that development should be relatively smooth sailing.

Thanks for you patience! And if you haven't checked out WordSwing in a while, please do! And don't forget to let me know what you think.


Learning Chinese characters the most common first

With thousands of Chinese characters to master, it can be daunting to get going and plough through them.

To help you out a bit, we would like to introduce WordSwing's frequency-ordered spaced repetition gadget:

SRS activity screenshot

This gadget will introduce the most frequent 870 Chinese characters in descending frequency order, that is, the most common ones first.

Learning characters in frequency order allows you to benefit from the very steep cumulative frequency curve intrinsic to the Chinese language. I'll illustrate with a plot from WordSwing's Character Knowledge Estimator:

SRS activity screenshot

This plot illustrates how the number of characters you know (horizontal axis) relates to the fraction of typical text that is comprised of those characters (vertical axis). This curve is drawn assuming that the probability you know a character is proportional to the frequency of the character. So in the plot above, if you know 870 characters then you can recognize the characters comprising 79% of typical text.

Now, if you learn the characters in strict frequency order then the curve is even slightly steeper. Thus, if you go through all 870 characters that are part of our new Frequency-ordered SRS gadget then you will actually know characters corresponding to 86% of typical text. In contrast, if you were to just pick 870 characters completely at random from the ~5000 or so most common characters, on average these would comprise only 9% of typical text. So it's really worth focusing on the common ones first.

By learning the 870 most common characters, you get 86% of typical text, with a random batch of 870 characters you would only get 9%!

Why yet another SRS tool?

Spaced repetition is one of the most powerful tools in the vocabulary learning toolbox. If you're not familiar with it, check out Hacking Chinese's introduction to SRS.

WordSwing's version of spaced repetition works much like the popular Anki spaced repetition program, although our version is not designed as a general purpose tool.

Instead, it's optimized to the task of learning Chinese characters. There are a number of features that make it particularly well-suited to character learning:

  1. As characters are introduced, common words that are comprised solely of already-introduced characters will be added to the mix. Adding words along the way allows you to practice the characters in the context of the words in which they are used, giving your brain more fodder for building associations.
  2. When you review a composite word, like 学生, if you indicate that you still remember it, then this also affects the review intervals for 学 and 生. In this way we leverage the intrinsic structure of the language to reduce the number of card views to get all vocabulary items to the highest review stage. However, if you indicate you don't remember 学生 then this doesn't affect the review intervals for 学 or 生 because it's not clear which part you forgot or whether you simply forgot that those two characters form this word.
  3. At each iteration, you explicitly choose whether you want to introduce a new character, review an item, or introduce a new word. This gives you full control over the balance of new material and keeping your review queue manageable.
  4. The gadget is wired up to the rest of WordSwing, so you have immediate access to related words, character decompositions, example sentences etc. By taking a moment to drill down and explore a word, you're giving your brain a chance to make additional associations that will help embed the new vocabulary item in your long-term memory.

We hope you enjoy our new gadget and, as always, welcome any and all feedback you have.

Have fun learning!

Page: 8 of 11

Learn a language by swinging up to ever higher levels of proficiency by effectively using the language you've learned so far.

Keep up to date