diff --git a/.gitignore b/.gitignore
index b5d6694..135bd5c 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,3 +4,4 @@ src/data/samples
 src/data/processing/
 prod/
 *.log
+src/data/unigram_freq.csv
diff --git a/src/data/database b/src/data/database
index c5ef65d..e69de29 100644
Binary files a/src/data/database and b/src/data/database differ
diff --git a/src/data/scripts/pipeline.sh b/src/data/scripts/pipeline.sh
index 2ebac2f..aac7f97 100755
--- a/src/data/scripts/pipeline.sh
+++ b/src/data/scripts/pipeline.sh
@@ -1,10 +1,10 @@
 #!/bin/bash
 cd ../processing
 
-jq -r '. | select((.pos=="noun") or (.pos=="verb") or (.pos=="adj") or (.pos=="adv")) | select((.word | test("[^a-z]"))|not) | {word: .word, type: .pos, pronunciation: .sounds[0].ipa, definitions: (try .senses|map({definition: .glosses|join(" "), form_of: (.form_of!=null)}))}' \
+jq -r '. | select((.pos=="noun") or (.pos=="verb") or (.pos=="adj") or (.pos=="adv")) | select((.word | test("[^a-z]"))|not) | {word: .word, type: .pos, pronunciation: .sounds[0].ipa, definitions: (try .senses|map({definition: .glosses|join(" "), form_of: (.form_of!=null) , plural: (try .tags catch []|any(.=="plural")) }))}' \
 wiktionary.json > wiktionary-p1.json
 
-jq --slurp '. | group_by(.word)[] | {word:.[0].word, pronunciation:.[0].pronunciation, meanings:([.[]|{type:.type, definitions:[try .definitions][]}] )}' \
+jq --slurp '. | group_by(.word)[] | {word:.[0].word, pronunciation:.[0].pronunciation, meanings:[[([.[]|{type:.type, definitions:[try .definitions][]}] | to_entries[] | [{type: .value.type, index: .key, definitions: .value.definitions}][] )] | group_by(.type) | map({type: .[0].type, index: .[0].index, definitions: [.[].definitions|select(length > 0)]}) | sort_by(.index)[] | del(.index) ] }' \
 wiktionary-p1.json > wiktionary-p2.json
 
 jq --slurp '.' wiktionary-p2.json > wiktionary-p3.json