Welcome!

Introduction

Let me give you some introduction here as the composer of the first wordset in Lojban at Memrise: My main and most important goal here was that I needed a way -- and also wanted a way for everyone else to easily and joyfully learn Lojban vocabulary. Memrise seems to provide this ability, so with the help of some others, the gismu wordset was born.

Contribute

How you can help with the current wordsets? Quite easily! As you may noticed, besides the basic information (Lojban word, English definition) I, for some of the words at the beginning, added additional information. These fall into the categories of:

  • Mnemonics (aka mems, these are little lines of English text which help you memorizing a given word),
  • audio (of the Lojban word),
  • etymology (taken from Wiktionary)
  • example of usage in a bridi,
  • easily confused with (due to morphological similarity or similar definitions).

Because of these require manual work, especially for mems, for which you also need to have a good idea, I require your help as well! If you see a word with no mem, or one which already has some but you feel you can do better, please feel free to add yours to it, and the votes will decide which version people like more.

It's similar case with examples.

Audio and etymology is very simple to add. You can copy etymology from Wiktionary, but please pay attention that Memrise's editor requires special kind of formatting. Take special care of how it handles new lines and how to make part of the text bold (for highlighting/emphasis). For audio, if you are confident in your Lojban pronunciation, which isn't that hard per se, and have a microphone, you can easily record a 1-2 second long sample of any word, and add it if it has no audio yet.

Future plans

This is just the beginning! In my vision for Lojban vocab learning in the near future, Memrise can play a key role. It's data structure allows us to upload all the vocab of Lojban, and organize them in intersecting wordsets. Think about cmavo for example: after all of the word-definition pairs are uploaded (now they are, only compunds are missing), wordsets can be easily created for subsets of cmavo, e.g. one for UI1, another for UI, another for all the evidentials, etc, and one for all the cmavo altogether! The great thing about this is, that if you start out with some kind of wordset, and want to broaden your vocab, Memrise won't forget that you had seen e.g. {.u'i} plenty of times in the UI1 wordset, and won't bug you with it that much from the beginning when you start out with the UI wordset.

I also made contact with the staff of Memrise, and it is hinted that if the specialty of Lojban requires special functions to be built for Memrise, they will be open to that in the future. I couldn't think of anything just yet. If you can, I'd be pleased to hear about it!

Questions, comments

Please, if you have questions, comments related to any (currently there is only one, but others are planned to come later, learn more above) Lojban wordsets, feel free to ask here, and if I get a notification from Disqus, I'm going to respond in a couple of days, or others might respond faster!

Updates

  • 2011.10.27: Cmavo wordset available in frequency order in all it's 538-word glory!

If the feature gets implemented later, I might update this post. If you see no changes, check to see whether I (or others) wrote in the comment section or the comments for the individual wordsets.

Posted by memrise_ 10/4/11, last update 8/17/12 (2 years ago)
  • This is a great idea :)I think that next word-list should be a a 50 most popular cmavo, regardles of selma'o.

    Posted by marzin 10/24/11 (2 years ago)
  • This is a great idea :)I think that next word-list should be a a 50 most popular cmavo, regardles of selma'o.

    Posted by marzin 10/24/11 (2 years ago)
  • Sad to see that the HTML formatting of my introductory text is discarded. And I can't even edit it. I might reformat it and paste to a comment later, so that I can keep it up to date.But what I'm happy about is that I see more and more people using the gismu wordset, and adding mems! Very good. :)

    Posted by Wizek 10/26/11 (2 years ago)
  • Sad to see that the HTML formatting of my introductory text is discarded. And I can't even edit it. I might reformat it and paste to a comment later, so that I can keep it up to date.But what I'm happy about is that I see more and more people using the gismu wordset, and adding mems! Very good. :)

    Posted by Wizek 10/26/11 (2 years ago)
  • I'd love to add cmavo too.The main problem with it currently is that I'm undecided as to how to extract it from the sources.I've written a little counter in Python which takes 2 kind of inputs:Source file full of lojban text.A csv dictionary of grammatical words which it looks for in the source text.Now, the first part of the problem is the following: (1) Cmavo are allowed to come in clusters, so we cannot rely on word separating whitespaces. One solution is to match at the beginning of every word, which would for example count .uisai and .uinai both as .ui Here comes the second problem: (2) nai and sai are also cmavos, therfore one cannot match them at the beginning of the word. This leads to the third part: (3) Think of the following text: ".ui .uicai .uicu'i .uinaibu'onai": How would you increment the counters at the end? The most basic and easiest solution would be: ui:4 meaning 4 instances of ui has been found, and nothing else. It's because we discard everything that comes afterwards of the first matched cmavo. This way none of the modifiers would be counted. this leads to the forth point of mine (4) find the possible cmavo clusters, then break them apart (we can actually do this quite easily) so that all are separated by spaces. Now count all the words. Here it is with the above example: ".ui .ui cai .ui cu'i .ui nai bu'o nai" ui:4, nai:2, cu'i:1, cai:1:bu'o:1. A lot better than before, but still has issues. A lot of the cmavos can be modified in a way that it seems like 2 or even 3 words' meaning is stuffed into one. Consider bu'o: It makes quite a lot of difference wheter it stands alone, coupled with cu'i or with nai. So I think the best statistics would take this into account. Considering the above example ".(ui) .(ui)(cai) .((ui)(cu'i)) .((ui)(nai))((bu'o)(nai))" resulting in ui:2, cai:1, cu'i:1, nai:2, bu'o:1, uinai:1, bu'onai:1, uicu'i:1 (but not including uicai as it only marks emphasys on the same scale).Now this is where my script is currently, and I couldn't really get it to give me to last kind of output. If you think you (as in anyone) can solve this problem, I'd be glad to hear from you. If you ask I can also give you the latest .py script.

    Posted by Wizek 10/26/11 (2 years ago)
  • I'd love to add cmavo too.The main problem with it currently is that I'm undecided as to how to extract it from the sources.I've written a little counter in Python which takes 2 kind of inputs:Source file full of lojban text.A csv dictionary of grammatical words which it looks for in the source text.Now, the first part of the problem is the following: (1) Cmavo are allowed to come in clusters, so we cannot rely on word separating whitespaces. One solution is to match at the beginning of every word, which would for example count .uisai and .uinai both as .ui Here comes the second problem: (2) nai and sai are also cmavos, therfore one cannot match them at the beginning of the word. This leads to the third part: (3) Think of the following text: ".ui .uicai .uicu'i .uinaibu'onai": How would you increment the counters at the end? The most basic and easiest solution would be: ui:4 meaning 4 instances of ui has been found, and nothing else. It's because we discard everything that comes afterwards of the first matched cmavo. This way none of the modifiers would be counted. this leads to the forth point of mine (4) find the possible cmavo clusters, then break them apart (we can actually do this quite easily) so that all are separated by spaces. Now count all the words. Here it is with the above example: ".ui .ui cai .ui cu'i .ui nai bu'o nai" ui:4, nai:2, cu'i:1, cai:1:bu'o:1. A lot better than before, but still has issues. A lot of the cmavos can be modified in a way that it seems like 2 or even 3 words' meaning is stuffed into one. Consider bu'o: It makes quite a lot of difference wheter it stands alone, coupled with cu'i or with nai. So I think the best statistics would take this into account. Considering the above example ".(ui) .(ui)(cai) .((ui)(cu'i)) .((ui)(nai))((bu'o)(nai))" resulting in ui:2, cai:1, cu'i:1, nai:2, bu'o:1, uinai:1, bu'onai:1, uicu'i:1 (but not including uicai as it only marks emphasys on the same scale).Now this is where my script is currently, and I couldn't really get it to give me to last kind of output. If you think you (as in anyone) can solve this problem, I'd be glad to hear from you. If you ask I can also give you the latest .py script.

    Posted by Wizek 10/26/11 (2 years ago)
  • And there you have the cmavo. :)http://www.memrise.com/set/100...It's not exactly in the order I described 6 hours earlier, and misses compound cmavo, but I thought having all of these (and in a nice order) is a lot better than having no cmavo here at all!

    Posted by Wizek 10/27/11 (2 years ago)
  • And there you have the cmavo. :)http://www.memrise.com/set/100...It's not exactly in the order I described 6 hours earlier, and misses compound cmavo, but I thought having all of these (and in a nice order) is a lot better than having no cmavo here at all!

    Posted by Wizek 10/27/11 (2 years ago)
  • I'm using this, and it's fantastic .ui.ki'e.I would like to suggest, though, that perhaps the selma'o should be omitted from the definitions, as in many cases it sort of gives away the answer.

    Posted by Pavitra 10/29/11 (2 years ago)
  • Cheers mate for this, Lojban (along with Icelandic, Finnish and Mandarin) is one of those languages I'd like to pick bits and pieces of up slowly, but perhaps that process can be sped up a bit this way - although the non-cmavo might have to wait till later, tavla I can remember, but 2-3 character mini-words are the ones that confuse me the most ;) Hopefully this helps develop interest in Lojban though, I'd love to see the casual learner community of the language grow (or "grow more visible") so I don't feel as lonely ;)

    Posted by ceige 11/16/11 (2 years ago)
  • Cheers mate for this, Lojban (along with Icelandic, Finnish and Mandarin) is one of those languages I'd like to pick bits and pieces of up slowly, but perhaps that process can be sped up a bit this way - although the non-cmavo might have to wait till later, tavla I can remember, but 2-3 character mini-words are the ones that confuse me the most ;) Hopefully this helps develop interest in Lojban though, I'd love to see the casual learner community of the language grow (or "grow more visible") so I don't feel as lonely ;)

    Posted by ceige 11/16/11 (2 years ago)
  • I'm glad you like it :)

    Posted by Wizek 11/16/11 (2 years ago)
  • I'm glad you like it :)

    Posted by Wizek 11/16/11 (2 years ago)
  • Rejoice those who dislike selma'o besides cmavo, as such wordlist has just been added!

    Posted by Wizek 1/24/12 (2 years ago)
  • Rejoice those who dislike selma'o besides cmavo, as such wordlist has just been added!

    Posted by Wizek 1/24/12 (2 years ago)
  • Here's a tiny helper app for registering etymology mems, which formats an etymology list copied from a wiktionary page for you: http://jsfiddle.net/ento/JGvTw/2/embedded/result/

    Currently it just takes care of correctly formatting bold text and newlines.

    Posted by ento 8/7/12 (2 years ago)
  • Quite nice!

    Are you aware that it doesn't work in firefox? Maybe because it doesn't support copy events?

    Posted by Wizek 8/9/12 (2 years ago)
  • Thanks for pointing out! I was lazy enough to support only my primary browser. This one works for Firefox (Mac), Chrome (Mac), Safari (Mac) and IE: http://jsfiddle.net/ento/AuDsR/embedded/result/

    Posted by ento 8/17/12 (2 years ago)

Recent threads

This forum doesn't have any recent activity