landonepps
commented
1 year ago Use https://github.com/ianki/MecabUnidic/releases/tag/v3.1.0 Download the latest unidic-csj-x.x.x.zip from https://clrd.ninjal.ac.jp/unidic/back_number.html#unidic_csj Copy char.bin, matrix.bin, sys.dic, and unk.dic from the downloaded archive to the support folder of MecabUnidic. Change dicrc to
; List of features
; f[0]: pos1
; f[1]: pos2
; f[2]: pos3
; f[3]: pos4
; f[4]: cType
; f[5]: cForm
; f[6]: lForm
; f[7]: lemma
; f[8]: orth
; f[9]: pron
; f[10]: orthBase
; f[11]: pronBase
; f[12]: goshu
; f[13]: iType
; f[14]: iForm
; f[15]: fType
; f[16]: fForm
; f[17]: iConType
; f[18]: fConType
; f[19]: type
; f[20]: kana
; f[21]: kanaBase
; f[22]: form
; f[23]: formBase
; f[24]: aType
; f[25]: aConType
; f[26]: aModType
; f[27]: lid
; f[28]: lemma_id
cost-factor = 700
bos-feature = BOS/EOS,*,*,*,*,*,*,*,*,*,*,*,*,*,*,*,*
eval-size = 10
unk-eval-size = 4
config-charset = utf8
node-format-unidic22 = %m\t%f[0],%f[1],%f[2],%f[3],%f[4],%f[5],%f[6],%f[7],%f[8],%f[9],%f[10],%f[11],%f[12],"%f[13]","%f[14]","%f[15]","%f[16]","%f[17]","%f[18]",%f[19],%f[20],%f[21],%f[22],%f[23],"%f[24]","%f[25]","%f[26]",%f[27],%f[28]\n
unk-format-unidic22 = %m\t%f[0],%f[1],%f[2],%f[3],%f[4],%f[5]\n
bos-format-unidic22 =
eos-format-unidic22 = EOS\n
node-format-verbose = surface:%m\tpos1:%f[0]\tpos2:%f[1]\tpos3:%f[2]\tpos4:%f[3]\tcType:%f[4]\tcForm:%f[5]\tlForm:%f[6]\tlemma:%f[7]\torth:%f[8]\tpron:%f[9]\torthBase:%f[10]\tpronBase:%f[11]\tgoshu:%f[12]\tiType:%f[13]\tiForm:%f[14]\tfType:%f[15]\tfForm:%f[16]\tiConType:%f[17]\tfConType:%f[18]\tlType:%f[19]\tkana:%f[20]\tkanaBase:%f[21]\tform:%f[22]\tformBase:%f[23]\taType:%f[24]\taConType:%f[25]\taModType:%f[26]\tlid:%f[27]\tlemma_id:%f[28]\n
unk-format-verbose = surface:%m\tpos1:%f[0]\tpos2:%f[1]\tpos3:%f[2]\tpos4:%f[3]\tcType:%f[4]\tcForm:%f[5]\n
bos-format-verbose =
eos-format-verbose = EOS\n
node-format-chamame = \t%m\t%f[9]\t%f[6]\t%f[7]\t%F-[0,1,2,3]\t%f[4]\t%f[5]\t%f[23]\t%f[12]\n\n
unk-format-chamame = \t%m\t\t\t%m\t未知語\t\t\t\t\n
bos-format-chamame = B
eos-format-chamame =
Is it possible to add more unidic dictionaries (the ones on the official website) to the written/spoken Japanese one? I don't know if adding other dictionaries, for detecting old speech for example, would mean that I'd have to select that dictionary in the settings or if it can be merged with the other files.