How Does Cloze Listening Add Content?

I realize that Listening is still in Beta, but I am curious about how and when new listening sentences are added.

I’m doing exclusively Italian. Sometimes my total sentence count goes up, and sometimes not. For the past week or two I’ve been stalled at 850 sentences, and even taking the “ready to review” count down to zero adds no new content. Whereas in the past, sometimes new sentences would appear if I exhausted the review queue. Surely it cannot be that there are only 850 sentences available and I have now burned through them all?


New feature - Cloze-Listening. Are you talking about this? It seems to have 1000 sentences. ( I can’t link it, but it is called New feature -Cloze-Listening from mike dated November 2020.


Correct. I set many to ignore that were just names in English, or similar - so maybe I truly have maxed it out. If so, that would be a disappointment, as there are thousands of sentences available in the source files.


Dcarl1! Well, it might be a disappointment for you to go through the whole Cloze Listening with all the authentic voices, but for me it sounds like a great achievement because I am still being babysat by Carla and Giorgio. I think mike might add more sentences for you if you ask. Good job and good luck!


That may be indeed be the case at the moment if you’ve ignored a bunch.

We should be able to add a lots more sentences. Thanks for letting us know about the names in English. Have you noticed any other issues? What have you thought about the sentence quality? Anything else you think should be added/changed before do add more sentences?


I did some recording/reviewing myself for the database in English, so I know how variable sound quality can be! I like that variability as it helps train my ear. In general 75% is quite good, 20% a bit tough to make out, and 5% unbearable/unclear. The ratio itself works for me, and I like being challenged.

I know others here have differing opinions on sentences that are purely or mostly just English names with no other native content (to make up an example: “Elvis Presley”). Some find these worthwhile for ear training - I do not, and “ignore” them.


@mike I have to say English names are no problem for me, they’re few and far between, they’re there for a reason and are light relief from some trickier sentences. It will be some time before I crash through the entire section so I’m more than OK with how things are. There are very few sentences I have needed to report or ignore so let’s hope you don’t go changing;-)


I discovered the feature today! It would be great to have it for Japanese.

The sentences look like those the Mozilla project collected, which should be used to train speech recognition.

I tried a few from the Italian set to check the quality, and I found a few that are really unrecognizable, due to noise, volume or pronunciation. But the concept is very useful, it would be great if native speakers could give feedback for various languages.