Hi! I was wondering what it would take to support Thai language word segmentation, for instance based on https://github.com/tlwg/swath
转发➕关注@彭大强同学 送3条价值108元的cike快充线！6月15日开奖，安卓苹果任选[赞]http:\/\/t.cn\/Ai98zfah \n\n苹果官方MFi认证，支持PD快充，30分钟能给手机充50%，线体铝合金抛光处理、1.5m编织线设计，好看又耐用，可承受80KG吊重，弯折30000+次……\n\n使用一段时间，真的很喜欢，好看又耐用[笑而不语]#尤物好物榜#
上面是例子
I want to use the custom score to inherit the original rating. What should I do?
the riot search
Hello :) Just found this project on GitHub... we are looking for a full text search engine for our photo app: https://github.com/photoprism/photoprism
Does riot also support numeric value and date range searches?
i tried to index from file which has size of 10M , when index finished , i found it cost 1G memory
i init riot engine with parameter Using: 1,
GseDict: "zh",
GseDict: "zh",
does anyone know why is cost so much memory ?
and how could i shrink it ?
怎么回事儿？
Not every image needs 1 GB :)
Oh sorry, thought this is a PhotoPrism question! In that case, I don't know...
text := "那里湖面总是澄清, 那里空气充满宁静"
seg := gse.New()
log.Debug().Interface("cut", seg.Cut(text, true)).Interface("cut all", seg.CutAll(text)).Msg("")
// output: cut=["那里","湖面","总是","澄清",", ","那里","空气","充满","宁静"] cut all=["那里","里湖","湖面","总是","澄清",","," ","那里","空气","充满","宁静"]
// seg.LoadDict("./dictionary.txt")
var tr idf.TextRanker
tr.WithGse(seg)
result := tr.TextRank(text, 5)
log.Debug().Interface("text rank", result).Msg("")
// output: text rank=[{},{},{}]
var te idf.TagExtracter
te.WithGse(seg)
if err := te.LoadIdf(); err != nil {
log.Error().Err(err).Msg("")
}
segments := te.ExtractTags(text, 5)
log.Debug().Interface("segments", segments).Msg("")
// output: segments=[{},{},{},{},{}]
TextTrank and ExtractTags not works. Can someone help me?
gse v0.64.1 with go 1.15.6
@lastzero I'm using Photoprism now. It's amazing. And from photoprism, I learned how to work with TF using golang. Thank you.