smartmachine 0.6.0 → 0.7.0

data/lib/smart_machine/grids/solr/config/configsets/sample_techproducts_configs/conf/lang/stopwords_ckb.txt

@@ -0,0 +1,136 @@
+# set of kurdish stopwords
+# note these have been normalized with our scheme (e represented with U+06D5, etc)
+# constructed from:
+# * Fig 5 of "Building A Test Collection For Sorani Kurdish" (Esmaili et al)
+# * "Sorani Kurdish: A Reference Grammar with selected readings" (Thackston)
+# * Corpus-based analysis of 77M word Sorani collection: wikipedia, news, blogs, etc
+
+# and
+و
+# which
+کە
+# of
+ی
+# made/did
+کرد
+# that/which
+ئەوەی
+# on/head
+سەر
+# two
+دوو
+# also
+هەروەها
+# from/that
+لەو
+# makes/does
+دەکات
+# some
+چەند
+# every
+هەر
+
+# demonstratives
+# that
+ئەو
+# this
+ئەم
+
+# personal pronouns
+# I
+من
+# we
+ئێمە
+# you
+تۆ
+# you
+ئێوە
+# he/she/it
+ئەو
+# they
+ئەوان
+
+# prepositions
+# to/with/by
+بە
+پێ
+# without
+بەبێ
+# along with/while/during
+بەدەم
+# in the opinion of
+بەلای
+# according to
+بەپێی
+# before
+بەرلە
+# in the direction of
+بەرەوی
+# in front of/toward
+بەرەوە
+# before/in the face of
+بەردەم
+# without
+بێ
+# except for
+بێجگە
+# for
+بۆ
+# on/in
+دە
+تێ
+# with
+دەگەڵ
+# after
+دوای
+# except for/aside from
+جگە
+# in/from
+لە
+لێ
+# in front of/before/because of
+لەبەر
+# between/among
+لەبەینی
+# concerning/about
+لەبابەت
+# concerning
+لەبارەی
+# instead of
+لەباتی
+# beside
+لەبن
+# instead of
+لەبرێتی
+# behind
+لەدەم
+# with/together with
+لەگەڵ
+# by
+لەلایەن
+# within
+لەناو
+# between/among
+لەنێو
+# for the sake of
+لەپێناوی
+# with respect to
+لەرەوی
+# by means of/for
+لەرێ
+# for the sake of
+لەرێگا
+# on/on top of/according to
+لەسەر
+# under
+لەژێر
+# between/among
+ناو
+# between/among
+نێوان
+# after
+پاش
+# before
+پێش
+# like
+وەک

data/lib/smart_machine/grids/solr/config/configsets/sample_techproducts_configs/conf/lang/stopwords_cz.txt

@@ -0,0 +1,172 @@
+a
+s
+k
+o
+i
+u
+v
+z
+dnes
+cz
+tímto
+budeš
+budem
+byli
+jseš
+můj
+svým
+ta
+tomto
+tohle
+tuto
+tyto
+jej
+zda
+proč
+máte
+tato
+kam
+tohoto
+kdo
+kteří
+mi
+nám
+tom
+tomuto
+mít
+nic
+proto
+kterou
+byla
+toho
+protože
+asi
+ho
+naši
+napište
+re
+což
+tím
+takže
+svých
+její
+svými
+jste
+aj
+tu
+tedy
+teto
+bylo
+kde
+ke
+pravé
+ji
+nad
+nejsou
+či
+pod
+téma
+mezi
+přes
+ty
+pak
+vám
+ani
+když
+však
+neg
+jsem
+tento
+článku
+články
+aby
+jsme
+před
+pta
+jejich
+byl
+ještě
+až
+bez
+také
+pouze
+první
+vaše
+která
+nás
+nový
+tipy
+pokud
+může
+strana
+jeho
+své
+jiné
+zprávy
+nové
+není
+vás
+jen
+podle
+zde
+už
+být
+více
+bude
+již
+než
+který
+by
+které
+co
+nebo
+ten
+tak
+má
+při
+od
+po
+jsou
+jak
+další
+ale
+si
+se
+ve
+to
+jako
+za
+zpět
+ze
+do
+pro
+je
+na
+atd
+atp
+jakmile
+přičemž
+já
+on
+ona
+ono
+oni
+ony
+my
+vy
+jí
+ji
+mě
+mne
+jemu
+tomu
+těm
+těmu
+němu
+němuž
+jehož
+jíž
+jelikož
+jež
+jakož
+načež

data/lib/smart_machine/grids/solr/config/configsets/sample_techproducts_configs/conf/lang/stopwords_da.txt

@@ -0,0 +1,110 @@
+ | From svn.tartarus.org/snowball/trunk/website/algorithms/danish/stop.txt
+ | This file is distributed under the BSD License.
+ | See http://snowball.tartarus.org/license.php
+ | Also see http://www.opensource.org/licenses/bsd-license.html
+ |  - Encoding was converted to UTF-8.
+ |  - This notice was added.
+ |
+ | NOTE: To use this file with StopFilterFactory, you must specify format="snowball"
+
+ | A Danish stop word list. Comments begin with vertical bar. Each stop
+ | word is at the start of a line.
+
+ | This is a ranked list (commonest to rarest) of stopwords derived from
+ | a large text sample.
+
+
+og           | and
+i            | in
+jeg          | I
+det          | that (dem. pronoun)/it (pers. pronoun)
+at           | that (in front of a sentence)/to (with infinitive)
+en           | a/an
+den          | it (pers. pronoun)/that (dem. pronoun)
+til          | to/at/for/until/against/by/of/into, more
+er           | present tense of "to be"
+som          | who, as
+på           | on/upon/in/on/at/to/after/of/with/for, on
+de           | they
+med          | with/by/in, along
+han          | he
+af           | of/by/from/off/for/in/with/on, off
+for          | at/for/to/from/by/of/ago, in front/before, because
+ikke         | not
+der          | who/which, there/those
+var          | past tense of "to be"
+mig          | me/myself
+sig          | oneself/himself/herself/itself/themselves
+men          | but
+et           | a/an/one, one (number), someone/somebody/one
+har          | present tense of "to have"
+om           | round/about/for/in/a, about/around/down, if
+vi           | we
+min          | my
+havde        | past tense of "to have"
+ham          | him
+hun          | she
+nu           | now
+over         | over/above/across/by/beyond/past/on/about, over/past
+da           | then, when/as/since
+fra          | from/off/since, off, since
+du           | you
+ud           | out
+sin          | his/her/its/one's
+dem          | them
+os           | us/ourselves
+op           | up
+man          | you/one
+hans         | his
+hvor         | where
+eller        | or
+hvad         | what
+skal         | must/shall etc.
+selv         | myself/youself/herself/ourselves etc., even
+her          | here
+alle         | all/everyone/everybody etc.
+vil          | will (verb)
+blev         | past tense of "to stay/to remain/to get/to become"
+kunne        | could
+ind          | in
+når          | when
+være         | present tense of "to be"
+dog          | however/yet/after all
+noget        | something
+ville        | would
+jo           | you know/you see (adv), yes
+deres        | their/theirs
+efter        | after/behind/according to/for/by/from, later/afterwards
+ned          | down
+skulle       | should
+denne        | this
+end          | than
+dette        | this
+mit          | my/mine
+også         | also
+under        | under/beneath/below/during, below/underneath
+have         | have
+dig          | you
+anden        | other
+hende        | her
+mine         | my
+alt          | everything
+meget        | much/very, plenty of
+sit          | his, her, its, one's
+sine         | his, her, its, one's
+vor          | our
+mod          | against
+disse        | these
+hvis         | if
+din          | your/yours
+nogle        | some
+hos          | by/at
+blive        | be/become
+mange        | many
+ad           | by/through
+bliver       | present tense of "to be/to become"
+hendes       | her/hers
+været        | be
+thi          | for (conj)
+jer          | you
+sådan        | such, like this/like that

data/lib/smart_machine/grids/solr/config/configsets/sample_techproducts_configs/conf/lang/stopwords_de.txt

@@ -0,0 +1,294 @@
+ | From svn.tartarus.org/snowball/trunk/website/algorithms/german/stop.txt
+ | This file is distributed under the BSD License.
+ | See http://snowball.tartarus.org/license.php
+ | Also see http://www.opensource.org/licenses/bsd-license.html
+ |  - Encoding was converted to UTF-8.
+ |  - This notice was added.
+ |
+ | NOTE: To use this file with StopFilterFactory, you must specify format="snowball"
+
+ | A German stop word list. Comments begin with vertical bar. Each stop
+ | word is at the start of a line.
+
+ | The number of forms in this list is reduced significantly by passing it
+ | through the German stemmer.
+
+
+aber           |  but
+
+alle           |  all
+allem
+allen
+aller
+alles
+
+als            |  than, as
+also           |  so
+am             |  an + dem
+an             |  at
+
+ander          |  other
+andere
+anderem
+anderen
+anderer
+anderes
+anderm
+andern
+anderr
+anders
+
+auch           |  also
+auf            |  on
+aus            |  out of
+bei            |  by
+bin            |  am
+bis            |  until
+bist           |  art
+da             |  there
+damit          |  with it
+dann           |  then
+
+der            |  the
+den
+des
+dem
+die
+das
+
+daß            |  that
+
+derselbe       |  the same
+derselben
+denselben
+desselben
+demselben
+dieselbe
+dieselben
+dasselbe
+
+dazu           |  to that
+
+dein           |  thy
+deine
+deinem
+deinen
+deiner
+deines
+
+denn           |  because
+
+derer          |  of those
+dessen         |  of him
+
+dich           |  thee
+dir            |  to thee
+du             |  thou
+
+dies           |  this
+diese
+diesem
+diesen
+dieser
+dieses
+
+
+doch           |  (several meanings)
+dort           |  (over) there
+
+
+durch          |  through
+
+ein            |  a
+eine
+einem
+einen
+einer
+eines
+
+einig          |  some
+einige
+einigem
+einigen
+einiger
+einiges
+
+einmal         |  once
+
+er             |  he
+ihn            |  him
+ihm            |  to him
+
+es             |  it
+etwas          |  something
+
+euer           |  your
+eure
+eurem
+euren
+eurer
+eures
+
+für            |  for
+gegen          |  towards
+gewesen        |  p.p. of sein
+hab            |  have
+habe           |  have
+haben          |  have
+hat            |  has
+hatte          |  had
+hatten         |  had
+hier           |  here
+hin            |  there
+hinter         |  behind
+
+ich            |  I
+mich           |  me
+mir            |  to me
+
+
+ihr            |  you, to her
+ihre
+ihrem
+ihren
+ihrer
+ihres
+euch           |  to you
+
+im             |  in + dem
+in             |  in
+indem          |  while
+ins            |  in + das
+ist            |  is
+
+jede           |  each, every
+jedem
+jeden
+jeder
+jedes
+
+jene           |  that
+jenem
+jenen
+jener
+jenes
+
+jetzt          |  now
+kann           |  can
+
+kein           |  no
+keine
+keinem
+keinen
+keiner
+keines
+
+können         |  can
+könnte         |  could
+machen         |  do
+man            |  one
+
+manche         |  some, many a
+manchem
+manchen
+mancher
+manches
+
+mein           |  my
+meine
+meinem
+meinen
+meiner
+meines
+
+mit            |  with
+muss           |  must
+musste         |  had to
+nach           |  to(wards)
+nicht          |  not
+nichts         |  nothing
+noch           |  still, yet
+nun            |  now
+nur            |  only
+ob             |  whether
+oder           |  or
+ohne           |  without
+sehr           |  very
+
+sein           |  his
+seine
+seinem
+seinen
+seiner
+seines
+
+selbst         |  self
+sich           |  herself
+
+sie            |  they, she
+ihnen          |  to them
+
+sind           |  are
+so             |  so
+
+solche         |  such
+solchem
+solchen
+solcher
+solches
+
+soll           |  shall
+sollte         |  should
+sondern        |  but
+sonst          |  else
+über           |  over
+um             |  about, around
+und            |  and
+
+uns            |  us
+unse
+unsem
+unsen
+unser
+unses
+
+unter          |  under
+viel           |  much
+vom            |  von + dem
+von            |  from
+vor            |  before
+während        |  while
+war            |  was
+waren          |  were
+warst          |  wast
+was            |  what
+weg            |  away, off
+weil           |  because
+weiter         |  further
+
+welche         |  which
+welchem
+welchen
+welcher
+welches
+
+wenn           |  when
+werde          |  will
+werden         |  will
+wie            |  how
+wieder         |  again
+will           |  want
+wir            |  we
+wird           |  will
+wirst          |  willst
+wo             |  where
+wollen         |  want
+wollte         |  wanted
+würde          |  would
+würden         |  would
+zu             |  to
+zum            |  zu + dem
+zur            |  zu + der
+zwar           |  indeed
+zwischen       |  between
+