[WIP] GloVe trial

configure.ac

@@ -1,5 +1,5 @@
 AC_PREREQ(2.59)
-AC_INIT([groonga-word2vec], 0.0.1, [[email protected]])
+AC_INIT([groonga-word2vec], 0.0.2, [[email protected]])
 
 AC_CONFIG_MACRO_DIR([m4])
 AC_CONFIG_SRCDIR([word2vec/word2vec.cpp])
@@ -109,6 +109,17 @@ AC_ARG_ENABLE(word2vec,
 )
 AM_CONDITIONAL(NOINST_WORD2VEC, test "x$enable_word2vec" = "xno")
 
+# GloVe
+
+AC_ARG_ENABLE(glove,
+  [ --disable-glove   Don't install GloVe binary],
+  [case "${enableval}" in
+     yes | no ) enable_glove="${enableval}" ;;
+     *) AC_MSG_ERROR(bad value ${enableval} for --disable-glove) ;;
+   esac]
+)
+AM_CONDITIONAL(NOINST_GLOVE, test "x$enable_glove" = "xno")
+
 GROONGA_REQUIRED_VERSION=4.0.3
 PKG_CHECK_MODULES([GROONGA], [groonga >= ${GROONGA_REQUIRED_VERSION}])
 
@@ -162,3 +173,9 @@ if test "x$enable_word2vec" = "xno"; then
 else
   echo "  Word2Vec: yes"
 fi
+echo
+if test "x$enable_glove" = "xno"; then
+  echo "  GloVe: $enable_glove"
+else
+  echo "  GloVe: yes"
+fi

test/suite/glove_train/min_count.actual

@@ -0,0 +1,32 @@
+plugin_register word2vec/word2vec
+[[0,0.0,0.0],true]
+table_create Tags TABLE_PAT_KEY ShortText
+[[0,0.0,0.0],true]
+table_create Entries TABLE_NO_KEY
+[[0,0.0,0.0],true]
+column_create Entries title COLUMN_SCALAR ShortText
+[[0,0.0,0.0],true]
+column_create Entries tag COLUMN_SCALAR Tags
+[[0,0.0,0.0],true]
+column_create Entries tags COLUMN_VECTOR Tags
+[[0,0.0,0.0],true]
+load --table Entries
+[
+{"title": "FulltextSearch", "tag": "Library", "tags": ["Groonga", "Rroonga"]},
+{"title": "Database", "tag": "Server", "tags": ["MySQL", "PostgreSQL"]}
+]
+[[0,0.0,0.0],2]
+dump_to_train_file Entries title,tag,tags
+[[0,0.0,0.0],true]
+glove_vocab_count --min_count 1
+[[0,0.0,0.0],true]
+#|w| [glove_vocab_count] vocab_count -input-file db/db_w2v.txt -output-file db/db_glv.vocab -min-count 1
+glove_cooccur
+[[0,0.0,0.0],true]
+#|w| [glove_cooccur] cooccur -input-file db/db_w2v.txt -output-file db/db_glv.cooccur -vocab-file db/db_glv.vocab -overflow-file db/db_glv.oveflow
+glove_shuffle
+[[0,0.0,0.0],true]
+#|w| [glove_shuffle] shuffle -input-file db/db_glv.cooccur -output-file db/db_glv.shuffle -temp-file db/db_glv.shuffle_temp
+glove_train
+[[0,0.0,0.0],true]
+#|w| [glove_shuffle] glove -input-file db/db_glv.shuffle -save-file db/db_glv -vocab-file db/db_glv.vocab -gradsq-file db/db_glv.gradsq

test/suite/glove_train/min_count.test

@@ -0,0 +1,20 @@
+plugin_register word2vec/word2vec
+
+table_create Tags TABLE_PAT_KEY ShortText
+
+table_create Entries TABLE_NO_KEY
+column_create Entries title COLUMN_SCALAR ShortText
+column_create Entries tag COLUMN_SCALAR Tags
+column_create Entries tags COLUMN_VECTOR Tags
+
+load --table Entries
+[
+{"title": "FulltextSearch", "tag": "Library", "tags": ["Groonga", "Rroonga"]},
+{"title": "Database", "tag": "Server", "tags": ["MySQL", "PostgreSQL"]}
+]
+
+dump_to_train_file Entries title,tag,tags
+glove_vocab_count --min_count 1
+glove_cooccur
+glove_shuffle
+glove_train

vendor/Makefile.am

@@ -14,3 +14,7 @@ if NOINST_WORD2VEC
 else
 bin_PROGRAMS += word2vec/word2vec
 endif
+if NOINST_GLOVE
+else
+bin_PROGRAMS += glove/cooccur glove/glove glove/shuffle glove/vocab_count
+endif