rbtagger 0.4.2 → 0.4.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/lib/brill/tagger.rb +28 -0
- data/lib/rbtagger/version.rb +1 -1
- metadata +1 -1
data/lib/brill/tagger.rb
CHANGED
@@ -26,6 +26,34 @@ module Brill
|
|
26
26
|
tag(text).select{|t| t.last.match(/NN/) }
|
27
27
|
end
|
28
28
|
|
29
|
+
# see: http://cpansearch.perl.org/src/ACOBURN/Lingua-EN-Tagger-0.15/Tagger.pm
|
30
|
+
def noun_phrases(text)
|
31
|
+
# ?:$PREP|$DET|$NUM)
|
32
|
+
#
|
33
|
+
tags = tag(text.gsub(/[^\w]/,' '))
|
34
|
+
phrases = []
|
35
|
+
phrase = []
|
36
|
+
mark = -1
|
37
|
+
|
38
|
+
tags.each_with_index do|tag,i|
|
39
|
+
if phrase.empty?
|
40
|
+
mark = i if tag.last.match(/PRP\$|DT/)
|
41
|
+
if tag.last == 'NNP' and mark != -1
|
42
|
+
phrase = [ tags[mark..i] ]
|
43
|
+
#mark = -1
|
44
|
+
end
|
45
|
+
mark = -1 if i - mark > 8
|
46
|
+
elsif tag.last.match(/NN/)
|
47
|
+
phrase << tag
|
48
|
+
else
|
49
|
+
phrases << phrase
|
50
|
+
phrase = []
|
51
|
+
mark = -1
|
52
|
+
end
|
53
|
+
end
|
54
|
+
phrases
|
55
|
+
end
|
56
|
+
|
29
57
|
# returns similar results as tag, but further reduced by only selecting nouns
|
30
58
|
def suggest( text, max = 10 )
|
31
59
|
tags = tag(text)
|
data/lib/rbtagger/version.rb
CHANGED