maybe a token filter that limit tokens by frequency is useful to minhash

[cclient]

feature:

filter tokens like

select count(*) as num,token from tokens group by token order by num desc limit 128;

then do minhash

[jimczi]

Can you describe the feature a bit more ? What do you mean by frequency ? The frequency of the terms inside the document ? If so I think it would be better to implement it directly inside the minhash filter since it seems to be only useful there ?

[colings86]

No further feedback. If you can provide more details on the feature request then please post them here and we can reopen the issue