Thanks! I went and found the word in the lesson, and played it. The sound played, with no error in my console. Must’ve been a temporary bug! 
1 Like
Question for Dutch experts!
@duaal @xKimka
This course is based on the “Dutch Lexicon Project”, according to the description of the course.
Do you have any idea how I could find the original database??? My Dutch is only rudimentary so I don’t know how to search for it online.
I am hoping that there might be a PDF or an Excel file or something like that online somewhere.
The reason I ask is that the very common words in any language - and Dutch is no exception - tend to have multiple meanings and multiple grammatical categories. In order to add the correct grammatical categories to a word like “om”, it would be good to know if “om” appears so early in this database because it was included with all its meanings, or does it appear there only because of more limited meanings?
I looked it up on the free vandale website and found this:
http://vandale.nl/gratis-woordenboek/nederlands-engels/vertaling/om#.WtPET-iuzrc
So now I am wondering how to improve the entry for “om” on the memrise course. Do I add all those meanings and categories? Or just the most frequent usage of “om”? And, if I want to do the latter, how do I find this information, if at all?
I hope my rambling post is halfway coherent. If not, please let me know!
Did you already look here? http://crr.ugent.be/programs-data/lexicon-projects
I don’t know if that’s what you need. It all looks too complicated for me. 
1 Like
That study seems to be about recognition rather than the meaning of the words. So I would expect all meanings to be grouped together.
It won’t be possible to include all possible usages of "om’ and keep it workable for learning. So what entry do you have in the course at the moment?
About frequency, have you tried
or
Those sites give some indication about frequency.
1 Like
yep. AFAIK this course is based on the SUBTLEX2 data, if I’m not mistaken. Besides, the file also contains data about Age of Acquisition (when kid’s learn words) and the word recognition speed (how fast readers get the meaning of a word). This course however (and the 1001-2000 course) seems to apply the SUBTLEX2 word frequency data that can also be found in this dataset:
☆ The Center for Reading Research; Department of Experimental Psychology of Ghent University : “The Dutch Lexicon Project 2” ☆ crr » The Dutch Lexicon Project 2 made available ☆ c.f. esp. file with SUBTLEX2 word frequency data ☆ http://crr.ugent.be/papers/DLP2datasetarticle_prevalence.xlsx
I used the SUBTLEX2 data for a quite similar course, that is quite useful for all user that have learned one of the “5000 most common words in Dutch” courses:
…check out the syllabus of this course.
1 Like
Thanks very much, I am going to do your courses next! I have already looked at the 5K course and also recommended it on the basis of what I saw. It looked very thorough and professional.
@amanda-norrsken I’ve seen this sort ofthing before where we get two possible answers that are both almost the same. I did not a screenshot of the previous example, but here’s one I saw now:
Really, 3 and 5 are the same answer; I only know that the “right” one is 5 because this was in the Learning phase and that’s the one they showed previously.
Is this an issue that can be fixed in the course, or it’s something with Memrise’s algorithm that can’t be avoided?
This is one of the biggest shortcomings of the platform itself, unfortunately.
1 Like
What to do:
-
Improve entries where possible. (obviously
) -
Users should delete donwloaded courses, log out, clear browser/app cache, log in again
-
Quite likely, this won’t do, sometimes phantom entries (old entries from previous phases of the course making process) remain. Creator should then run this sript: [Userscript] Attempt to Eliminate Phantom Entries It forces Memrise to refresh the course. This time Memrise should only use actual entries.
1 Like
Ah, the synonym problem … 
This is one of the most difficult problems to solve on the memrise platform. As Robert Alexander says, improving the definition is the best thing to do and I have actually already done so. Here is a screenshot of the database entry when I type in “sweetheart” as a search term:
As you can see, there are three entries that can be translated as “sweetheart”, “love” etc. I have added “not X, Y” to all of them, so what you are seeing must be one of those “phantom entries” that Robert mentions. Or you need to refresh your version of the course. I added these amendments some time in the last six weeks or so, I would say.
I hope this helps.
1 Like
I forgot to mention once again that this was on the web and not on the app. I doubt caching was an issue as @Robert-Alexander mentioned, unless Memrise downloaded the question and answers weeks ago, if you fixed it 6 weeks ago? That would seem an odd way to do it to me, but I suppose it’s possible. Oh well, I’m not blaming you, @amanda-norrsken, and I greatly appreciate your efforts in cleaning up the course! Alsjeblieft!!
2 Likes
I am sorry you are having this problem, Richard! As you can see, I have tried hard to eliminate it, so I can only conclude that it might have something to do with “phantom entries”, which I have read about elsewhere on the forum. I know that - during reviews - I have seen really old entries on the Swedish course I have been curating for over three years now. When I see them, I think, “How can that possibly still be floating around??? I am sure I changed that years ago???”
What do you think could be causing this discrepancy? Is it some kind of bug? Should we put this problem elsewhere, in the bugs section? Or alert someone from the memrise team, like @BeaTrisy or @MemriseSupport?
Memrise has never reacted to people reporting this database problem. Run neoncube’s userscript like Robert-Alexander already mentioned.
1 Like
I agree with @duaal; try the userscript if you want. Otherwise, don’t sweat it, I can live with the occasional “forced” errors, it just makes me practice that word or phrase a little bit more!
I have absolutely no clue how to go about doing this! Would they even work on a Chromebook, I wonder???
You’ll have to install Tampermonkey for Chrome first.
With Tampermonkey you can run the script in your browser. It may take a while to finish but the old database entries should be gone afterwards.
2 Likes
@amanda-norrskenThanks for being willing to give that a shot! I’m sure you’ll be able to! 
I have another problem to report. The word “eindelijk” appears twice in the same lesson, lesson 26:
This is on the web, by the way.
1 Like
Yeah, I’ve noticed that, too. I can only presume it was a mistake and I was wondering whether to just delete one of the entries?
On the other hand (this course is not “my” course, so I wonder about doing stuff like that 
), seeing as the two entries are identical, then there is no problem getting anything wrong when you review something.
Thanks so much! It would be brilliant if those old entries disappeared!
this course is not “my” course, so I wonder about doing stuff like that
I see it as maintenance, or fixing an overlooked error. Of course, it’s not that important in the grand scheme of things, so don’t do it if you’re not comfortable doing so!
1 Like