From 30d1bd804d1e9891287289388015bfecc5e25a12 Mon Sep 17 00:00:00 2001 From: Patrick Simianer Date: Tue, 7 Jun 2016 17:10:19 +0200 Subject: phrase2_extraction: adjacent nt, MAX_SRC_SZ --- phrase2_extraction/phrase2_extraction.rb | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/phrase2_extraction/phrase2_extraction.rb b/phrase2_extraction/phrase2_extraction.rb index 01bdae9..1f268cd 100755 --- a/phrase2_extraction/phrase2_extraction.rb +++ b/phrase2_extraction/phrase2_extraction.rb @@ -7,7 +7,7 @@ module PhrasePhraseExtraction DEBUG = false MAX_NT = 1 # Chiang: 2 MAX_SEED_NUM_WORDS = 3 # Chiang: 10 words, -> phrases! -MAX_SRC_SZ = 7 # Chiang: 5 words, -> words! +MAX_SRC_SZ = 8 # Chiang: 5 words, -> words! FORBID_SRC_ADJACENT_SRC_NT = true # Chiang:true class Rule @@ -649,6 +649,8 @@ def PhrasePhraseExtraction.remove_adjacent_nt rules prev = false end } + b +=begin c = false prev = false r.target.each { |i| @@ -663,6 +665,7 @@ def PhrasePhraseExtraction.remove_adjacent_nt rules end } b || c +=end } end -- cgit v1.2.3