In “Yajiuma no Mori”, we will tell you about a wide range of topics that are not limited to news and reviews.
If you search for “Twitter” with the following query, you can see your own tweets that were not judged safe.
from:自分のID -filter:safe
In short, it seems that the tweets that are caught in this search are things that “Twitter” has judged to be “Akanyaro”.I also tried it on the official account of Mado no Mori (@madonomori)But–
It seems that there are many cases where it is judged as bad simply by partial match, for example, “fool” (idiot) included in the title “I just did it” and “chichi” (milk) in “Patch Tuesday” are caught. (I’ve often heard that words like “milk” are also matched). It’s really pitiful that even “Embarcadero”, which provides app development tools, will be caught…
In addition, it seems that words such as “marathon”, “cafe latte” and “Ferrari” are also judged as out remarks. AI is booming these days, and I felt that it might be possible to create a smarter system by making use of it.