I use SkELL quite often, so I was glad to read that there has been an update to the example sentences that get shown. This was an occasional, irritating issue because the kinds of SkELL-assisted activities I usually do with my students are hampered by spelling mistakes and such. For them, the learners, the cleaner data should be beneficial.
The update was announced by James Thomas on the CorpusCALL FB group, the text is reproduced in italics here:
“SKELL If you are a user of SKELL, you might have noticed a recent improvement in the quality of the example sentences. This is thanks to the deletion of sentences that contained spelling mistakes and hapax legomena. While both of these things can be of interest, it is better that the 40 example sentences of a word or phrase are as accurate as possible.
There are 10,370 instances of the word ‘dolphin’, for example, in the full corpus. The algorithm that chooses the best 40 for learners now works with cleaner data.
It’s a nice improvement. Thanks, Vit.”
**Link to SkELL**
**Earlier posts on using SkELL**