Vocabulary dictionary

Kanji dictionary

Grammar dictionary

Sentence lookup

test
 

Forums - [Discussion] Better way to handle usage note validation

Top > renshuu.org > Feature Requests/Improvements

Page: 2 of 3



avatar
Seycalia
Level: 188

The “gaming” aspect does seem to complicate things. Honestly i’d be happy to look at usage notes without any reward, and i’m sure enough people would be willing to as well. Is it not possible to just not give any rewards for it? Removes any incentive to misuse the system as well. Or maybe a tiny reward, but really not enough to make it even worth misusing?

8
5 days ago
Report Content
avatar

True, the users we want probably aren’t going to find the proposed rewards to be much of an incentive anyway.

7
5 days ago
Report Content
avatar

I don't know exactly what the user distribution is and most of us are probably English speakers, but leaving the rating mostly to the community theoretically opens up the possibility for multi-language mnemonics, as in, mnemonics that are written in languages other than English.

If you speak English with a heavy accent it impacts what English words actually sound similar to a vocab term in your mind.

The question would just be how to handle those with mixed reception, since "just let the mike deal with it" won't work for 22 non-EN/JP languages. Plus, again, I don't know how many people would use these and how heavy the implication for dev time would be.

Also there'd need to be at least some guideline about what is and isn't to be considered appropriate, since you'd let people from various cultural backgrounds vote on that with limited supervision.

Edit: I didn't use this feature much so far and legit just now realised that mnemonics already have a language selection feature. I'm guessing I only ever sawEnglish ones because that's my display language outside the dictionary.

1
5 days ago
Report Content
avatar
🍏
Level: 323

Frankly, the most attractive point for me to use this app is the mnemonics. However, just as some of you have said, sometimes I just can't get it. I suppose it may due to my mother tongue is not English. At the same time, my mnemonics may not work for the others. Basically, for me, I just hide those which didn't work for me.kao_sorry.png

Concerning NSFW, I haven't seen much in this app (maybe they were already banned by Micheal?). Some said those kinds of mnemonics are impressive and easy to recall while personally I would rather keep them private (or friends only?) just to keep the platform as SFW as possible. I note that the defaulted setting of usage note is "public", may I suggest to default all notes as "private"? So, in case someone wants to share their mnemonics, they need to actively change the setting. In addition, tags like "NSFW", "slang", etc. can be used to "warn" the viewers about the contents if necessary.kao_fire.png

Concerning the issue on "objectively wrong", for me, there's no right or wrong in mnemonics as long as it can help me ringing the bell. As in the case of , the meanings of individual radials, for me, are somehow not important. I just take the advance of their appearances to memorize the words. However, for usages and meanings of a language, though it'll evolve over time, as a beginner, I'd rather stick to the traditional/ classical one. So, I agree with the idea on the voting system. While what I am thinking is, instead of voting "up/down/skip", "agree/disagree/questionable" can be used. During the process, viewers can justify whether the particular notes are acceptable or not. It's may be a good chance to learn more on those controversial notes.kao_punch.png

I think that's all I would love to say. The last but not the least, Merry Christmas to you all!kao_heart.png


5
5 days ago
Report Content
avatar
Tsuta13
Level: 340

Many good points have been made already. I think opt-in widget with yes/no/skip would be good. I would add, that a mnemonic that gets skiped by many should probably go Mike. Because I think that shows that it's not straightforward (similar to mixed up/down votes).


As for rewards - I find the system of word tagging good. Why not reuse that :)

5
5 days ago
Report Content
avatar
マイコー
Level: 283

Morning all! A few updates based on recent comments:
1. I am not sold on the reward factor - just putting out all the ideas I have.

2. Objectively, outside of something that is off-topic (not a mnemonic, or for example, a mnemonic for a hiragana character that just describes that shape, but that description has nothing to do with the reading of the character), or one that uses the wrong pronunciation altogether. Sometimes, they are close enough (almost like a "near-rhyme") that I let them pass, but let's say the word is こ and they are using こう in their mnemonic - that's going to, as someone pointed out, possible reinforce the wrong pronunciation.

I do love, though, when people do "combo" mnemonics - it's not where you have mnemonics for two words, but where they use another, similar Japanese word in the mnemonic. I do not have one to reference, though, sorry! Let's make one up: say the mnemonic is for a word けんき, which means "shopping". The mnemonic might say "John loves shopping: he's げんき for けんき." I think those are generally more memorable than not.

But if the "data" in the mnemonic is wrong, then I'd consider it an objective (and not subjective). Mnemonics are used when you *don't* know the word, so the learner is in the worst position possible to judge if the mnemonic is teaching them the "wrong" thing.

3. The reason you don't see nsfw mnemonics are because I've already filtered them out. Sometimes it's language, sometimes they are racist/sexist/(...ist), sometimes adult in nature. I'm sure you can guess that I see a few mnemonics a week that use the reading of ふく to represent a particular English word. They can be very catchy, just not a good fit for renshuu.


It's also worth noting that mnemonics are a HUGE draw for new users on renshuu - they are mentioned all the time in app reviews. I believe it's not just the presence, but the quality of them. Sure, we could open it up to where more of them make it through (that wouldn't if I was doing it myself) and leave it to users to block them or report them, but that will result in a non-zero drop in perceived value of the mnemonics and renshuu as a whole by new users. You are all (presumably) already invested in the platform, so you are fortunately willing to maybe deal with more "noise" because you already have an established value in being here.

12
4 days ago
Report Content
avatar

I think this would be a fantastic idea.

As I’m a very visual learner, I find mnemonics to be very useful (that might sound weird as mnemonics are words, but I’m also a visual thinker ((if that makes sense)) so they’re almost essential for some of the more difficult words). Quite often when I’m struggling to remember a word, I’ll check the mnemonics and be disappointed that they’re filled with very similar examples or ones that don’t make a lot of sense sometimes. I realise that might just be me, but some of them can be quite stupid and take up a spot that could be filled by something more helpful.

I’ll also echo the sentiment a couple of people have already made that I don’t think a reward should be necessary. I feel like the kind of people who would be motivated by that would be more likely to act disingenuously just to get the rewards. It might sound cynical but speaking as someone with some experience in reward based customer loyalty programs (in retail) it will always happen. Maybe not in the majority of cases but enough to muddy the waters. I realise also that not offering a reward may possibly lessen the number of people who participate. Maybe if it’s something like a small garden item or a rare gatcha for the top contributors that month? If it’s XP though it will definitely be farmed.

Anyway, those are my thoughts! Sorry I rambled a bit, but the mnemonics have always been important to me. Thanks for all your hard work Michael, this app is great :)

TL;DR: Firehose widget is a great idea, but imo rewards aren’t needed or at least shouldn’t be farmable.

5
4 days ago
Report Content
avatar
ハシュミナ
Level: 755

I can't speak for any reward motivation since I'm sitting on a mountain of Kao garden coins and eagerly waiting every year for the gacha coin event to drop so I can finally get some new ones so those are not really a motivation for me. If I had to suggest something, something like a profile ornament (similar to like the level-able hanko stamps). But I guess that would require new implementation so maybe just stick to what's already there.

For rewards in coin/XP form maybe to prevent farming put out a reasonable daily limit.

3
4 days ago
Report Content
avatar
Shamugan
Level: 464

I did not think I would write so soon again but I can't resist on that subject.

In my case, I would open the floodgate but with some statistical "security".
Like:
- Each notes should be validated (or pre-validated) by X user. Or even better X% of the active user (or advanced user, senior, etc, it's possible to add other criteria) that are actively and currently validating the notes. It would a first security (even with a positive ratio of thumb up, no validation).
- Second, enough thumbs up (or better a positive enough ratio of the thumbs up/down). The mixed ones could be check manually but I would do it with the second system after (perhaps a flag by the trusted user?).

But otherwise, I would not bother with the notes (mnemonics or not) that doesn't meet those criterion. It would probably not be perfect and may even prevent some good notes to be validated at first but with enough people involved, it will statically correspond to what the community like and found useful as a whole.
It's still possible to put a small reward for everyone. Or some tier reward for the people that really help the community, I'm thinking about the people that will maybe validate thousand of notes... not sure the "cheaters" would do that. The smallest should not be interesting enough for the cheater and the biggest one, too difficult to do. And there would be no need to check for people who incorrectly judge the notes, because with enough honest people (and I think in this community, there is enough honest people), those just won't be validated.

On a second part, a more logical system with a flag system for the rule breaker ones at any moment (nswf, inappropriate, etc) that immediately hide the note for example. And it's possible to add manual check for some of those (like the polemic ones). It could be also improve with time with a flag for particular case (a flag "other problem", "NA", etc). Then it could become another flag or another rule to let those pass.
Also a security/warning for people who validate too often some note that just break the rule. I think it's better that way because some notes are sometimes really hard to judge (the mnemonics are hard but the others are also complicated too sometimes). Except for the nswf ones and probably a few other cases. Those, with your choice for the community, could be shut down immediately with enough trusted/senior user or with enough flag from different trusted user. As for the more complicated/polemic -> manual check by you -> new flag to automate the system (like the automatic shut down flag) or new rule to make it easier to judge -> less work for you in the future.

Anyway just a few idea, hope it will be useful.

1
4 days ago
Report Content
avatar

I would personally like to see more word mnemonics rather than fewer. I can understand some people being irritated that there are "too many" for a single word, or too many that they don't find helpful, or too like another (which can happen as a previous mnemonic gets refined/improved upon by other users later to be "more helpful"). I myself am frequently mystified why some seem to get so many likes, but they still seem popular. But I feel that we have the benefit of being able to hide ones from our view that we don't want to see.


If a word was limited, say, to 5 mnemonics, I may not personally find any of the 5 personally helpful, and may miss being able to see one I COULD find helpful because it was written after the word reached a 5 (or whatever #) mnemonic limit and so cannot be shown. If I think there are too many, and only want to see 5, I can pick the 5 I like and hide the rest. Then, if another shows up that I find more helpful, I can pick the least helpful of the 5 I had still visible and hide that instead of missing out on the new one. Just as an example.


I can understand not wanting the list to grow too long, however, since it could get overwhelming and/or difficult to slog through. But I don't know what a good cutoff number could be. I would hope that within, say, 20 usage notes, there would exist enough helpful notes that users aren't really as inspired to create more, and it could be rather self-regulating in that regard. Though I don't know if that would necessarily be the case.


As for guidelines about "objectively bad" mnemonics, especially based on pronunciation "mistakes", I have seen plenty that seem to be so far off that the kana in the word seem to have been outright misread by the user writing the mnemonic. But I'm concerned with how strict the pronunciation matchup would need to be to not be rejected. Given the previous example from this thread, if I'm understanding it correctly, using a word with a こ sound for one with a こう sound, or vice versa, being considered "misleading", mnemonics could be impossible to have pass. But maybe I misunderstood that.


There are a lot of words in Japanese that have sound combinations that just don't have exact matches in English, or even sounds that are so rare in English anyway that one must get creative. Say someone used "at a race" in a mnemonic for しい . Sure, there may be a better grouping of words for a closer *actual* pronunciation, but this selection might still actually be helpful to other users. But also, an "exact" pronunciation like "at a rash she" could be super, super difficult to write something sensible enough to be memorable. This kind of pronunciation criterion could make just about any mnemonic "objectively bad" to many, let's say, more "purist" users. Also, a user might vote it down because they find it less helpful than other users based on the way their dialect of English (or the one that they're familiar with) pronounces different English words.


What would be too "misleading" of a pronunciation? For instance, I have one where I use "time many" for たまに . Is it "objectively bad" or "misleading" because the pronunciation is more like たいまに ? I do often find I get a bit creative with pronunciation representations, usually something that keeps popping into my head whenever I encounter the word in review before I form a mnemonic, but I always include the actual pronunciation in both hiragana and romaji (for newer users who may struggle with correctly reading all hiragana at a glance, so there's less chance of misunderstanding). **Side note: I generally "hear" the correct pronunciation (especially if there's a sound recording attached to the word) in my head when I *think* of the English words I use to remember it... it's possible that many other users may do the same, rather than getting stuck on the sounds of the actual English words, as seems to be assumed.**


I would like to see the bulk of restrictions, instead, being for NSFW, offensive, etc. But even there, I'm not sure how much restriction should be required, or has/hasn't been until now. For instance, I have seen mnemonics that included "whores" ホウ and "cocaine" こく, which might bother some people more than others. And referring back to たまに , the most popular mnemonic involves *potential* non-consentual touching of a woman's body (in what could be considered a fairly intimate area), which might bother, say, women who have experienced being touched without their consent. (I know I wouldn't want someone occasionally touching MY knees... though they'd probably only make that mistake once!) But then again, it is the most popular mnemonic for the word with 68 likes, so who's to say?


I do agree with a neutral "skip" option for judging, rather than simple pass or fail. But I'm not sure how neutral votes should be counted. Though I think if a mnemonic is getting more neutrals than fails, it should probably be passed along to being visible as it may not be explicitly helpful to the majority, but it's apparently not terribly harmful and could potentially get more "likes" from a broader audience. If that's the right way to put it.


As for rewards, I think I agree that rewards may lead to abuses from less scrupulous users fishing for rewards. But I'm open to options, though I think exp would be most useful to me, personally, (at least for the time being) since I have more coins than I can spend and am still working on getting different combinations of Kaos.


That's all I can think of to contribute, for now. Thank you for letting me contribute. And I hope everyone has a great holiday!

5
3 days ago
Report Content
avatar
Jules.33
Level: 37

First, I had no idea you were manually checking every one of those usage notes. I’m a relatively new learner, and now I’m embarrassed because of how many new usage notes I added already🙈. I really hope you didn’t have to re-review every note that I edited after saving. Like you mentioned, the usage feature is one of my favorite features, and I find it very helpful to both create my own mnemonics and read other people’s. I actually thought that the heart system was already being used for the purpose of identifying the best usage notes and determining which stayed public. I didn’t realize until reading this thread that I could hide the ones that I’m not interested in. I just favorited the ones I liked, even if there were multiple ones. After reading through the thread, here is my two cents:

1 . I see no point in the reward system. It just seems like an added feature that isn’t needed and that just complicates things. To be honest, as a new user, I don’t really find any of the current reward systems to be a motivating factor for me as to whether I practice or not.
2. I like the idea of the flagging system for inappropriate content. You could have a drop down menu with different colored flags, like a red flag gets the note immediately removed, a yellow flag needs to be reviewed further, and a green flag indicates that it’s a good note, but needs to be tweaked.
3. You already have a voting system in place with hearts. Can you use those to help determine which notes remain public and which get switched to private? Perhaps a grace period for new notes. After a certain amount of days, if a note doesn’t get any hearts then it gets dropped off the public list. But it would be a shame, if somebody was using that mnemonic and it disappeared from their list. Is it possible for a public note to become private for other users who have previously favorited it, so that they don’t lose it?
4. I agree that the default should be private and the person has to make it public.

With the flagging system, the hearts, and the grace period, I would think that the whole process could be mainly outsourced to the community. There could be an automatic message that gets sent to the person’s email when a note is flagged indicating that their note was either removed or switched to private, reminding them of the community guidelines, and giving them the option to edit it and send it back for further review if they choose. If individuals are repeatedly getting flagged for inappropriate content then they can lose their public privileges.


2
3 days ago
Report Content
avatar
モニカ7
Level: 1013

I agree with a lot of the suggestions that have been posted here. Personally as a mother of five kids who are all using renshuu I definitely want there to be a way to stop inappropriate mnemonics from getting out there before little eyes see them. I will happily volunteer to help with whatever system ends up being implemented. kao_think.png

2
3 days ago
Report Content
avatar
マイコー
Level: 283

Some notes on rarer words are the most valuable, but simply might not get the same coverage. Most things on the internet follow a 1% pattern (or something like that)

Of 100 users, 1% are logged in (doesn't exactly apply to renshuu)

Of those logged in, 1% will interact with something, like leaving a heart. You see the same thing on reddit - subreddits with huge numbers of followers (plus even more guests), with so few upvotes. It's not a bad thing, per se, but renshuu is pretty small, so when you go down the division rabbit hole, you're going to end up with many less "active" users for any given feature than you might expect.


If all works out, I'm going to make a test widget sometime soon. I might not act on the data, but I'd like to see how much data I get, and how quickly. If this system helped approve/disapprove even 20 mnemonics a day, that would be a huge dent in my workload.

6
3 days ago
Report Content
avatar
Jules.33
Level: 37

I think just changing the default to private might make a dent. I wonder how many people are not even aware of, or don’t bother with the drop-down at all. I bet a lot of people are just putting those usage notes in there without any consideration for whether they’re private or public. That little bit of extra effort needed to make the switch to public might make a big difference in the amount of notes you are getting. And as far as testing, that would be an easy change that you could see the results of right away.

1
3 days ago
Report Content
avatar

Some notes on rarer words are the most valuable, but simply might not get the same coverage.


This is precisely why I wouldn't want to see a policy along the lines of "get a like within X# of days or it disappears from view". I do branch out into words well outside of those presented in renshuu lessons, and they may not be "discovered" by users very frequently. Also, not everyone who finds a mnemonic useful necessarily gives it a "like", but probably wouldn't want to see it disappear either.


I have written mnemonics for a number of words I've come across that had no mnemonic already, and I don't know when they'll be seen, if at all (for some less common words). I hope they stay, in case they can help someone when eventually stumbled upon.


**Edit: I have made a number of my mnemonics/notes private... sometimes because they're mainly rewrites of existing ones with tweaks I find helpful for myself, but would likely be redundant to others, or for other reasons. Default privating mnemonics/other usage notes may help in reducing the number to review, but may also inadvertantly deprive the community of helpful notes if users are unaware they need to manually make them public or forget to select the public option. IMO. **

4
3 days ago
Report Content
avatar

mnemonics … written in languages other than English

いました。

Usage notes do have a language tag, although it wouldn’t surprise me if most contributors ignore it, and there’s no way to set it for more than one language.
The firehose widget should only show notes to mods if they are written a language that they speak. That will require mods to tell renshuu which languages they speak.

Mods might want to just fix the language tag if confronted by a note written in a language they don’t speak (or speak very well) and send it back into the queue.

2
3 days ago
Report Content
avatar
Shamugan
Level: 464

Some notes on rarer words are the most valuable, but simply might not get the same coverage. Most things on the internet follow a 1% pattern (or something like that)

Of 100 users, 1% are logged in (doesn't exactly apply to renshuu)

Of those logged in, 1% will interact with something, like leaving a heart. You see the same thing on reddit - subreddits with huge numbers of followers (plus even more guests), with so few upvotes. It's not a bad thing, per se, but renshuu is pretty small, so when you go down the division rabbit hole, you're going to end up with many less "active" users for any given feature than you might expect.


If all works out, I'm going to make a test widget sometime soon. I might not act on the data, but I'd like to see how much data I get, and how quickly. If this system helped approve/disapprove even 20 mnemonics a day, that would be a huge dent in my workload.

Yep but at the same time, it's "normal" without mechanisms to develop that participation.
And the deadlock with the notes doesn't help. Like "I'm sharing my notes with the community but it doesn't get publish/I don't get any suki point". I'm talking a bit egocentrically here but "I'm doing something for 'free' for the community and I don't even know if it was useful or not for other, so why should I continue?". It's a stupid basic human feeling but it feels nice when you know that what you're doing is useful for the community at a bare basic minimum. Feedback, at least, is really important for that. And when the active user don't have even that (or enough of that), they become inactive. Usually, in every community, it's always the same. New member are highly motivated but after some time, or past a certain point, the ratio of effort/reward is not enough anymore (member put also more effort with time, that's also difficult to have a good system for that). And by reward, I mean at least being able to know if what you're doing has any meaning to the community. If not, they became inactive and replaced by the next wave (so it's stay at that 1%).


I think reshuu's community is a bit stuck here but has the potential of not following that 1% rule (same for other aspect of the app like the community lists). Anyway, not saying it should go in that direction (for a number of different reasons, notably, it's a lot of work) but I believe it's possible in the case of that community

kao_great.png

.

1
3 days ago
Report Content
avatar
マイコー
Level: 283

Mis-tagged languages is maybe 1%, so extremely rare.

1
2 days ago
Report Content
avatar

Another obvious nit that needs to be addressed is that no one should be put in the position of approving their own work, and ideally, their friends either.

1
2 days ago
Report Content
avatar
マイコー
Level: 283

You wouldn't see your own, and you wouldn't see the names of other people in the firehose, so I think we're probably safe there.

2
2 days ago
Report Content
Getting the posts


Page: 2 of 3



Top > renshuu.org > Feature Requests/Improvements


Loading the list
Lv.

Sorry, there was an error on renshuu! If it's OK, please describe what you were doing. This will help us fix the issue.

Characters to show:





Use your mouse or finger to write characters in the box.
■ Katakana ■ Hiragana