aboutsummaryrefslogtreecommitdiffstats
path: root/scalding/src/main
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2018-08-27 16:40:15 -0700
committerBryan Newbold <bnewbold@archive.org>2018-08-27 16:40:15 -0700
commitb4eac17049e19d33b1a55664a7258c0f62f8a8c7 (patch)
treee71de0eb969f53765e4f7b8f5053b5ece3b28781 /scalding/src/main
parent309f40b66d474f12c0cfe60c449d43ae4bacb912 (diff)
downloadsandcrawler-b4eac17049e19d33b1a55664a7258c0f62f8a8c7.tar.gz
sandcrawler-b4eac17049e19d33b1a55664a7258c0f62f8a8c7.zip
make similarity score case-insensitive
Diffstat (limited to 'scalding/src/main')
-rw-r--r--scalding/src/main/scala/sandcrawler/Scorable.scala2
1 files changed, 1 insertions, 1 deletions
diff --git a/scalding/src/main/scala/sandcrawler/Scorable.scala b/scalding/src/main/scala/sandcrawler/Scorable.scala
index 5aac032..f7eb95d 100644
--- a/scalding/src/main/scala/sandcrawler/Scorable.scala
+++ b/scalding/src/main/scala/sandcrawler/Scorable.scala
@@ -72,7 +72,7 @@ object Scorable {
getStringOption(json2, "title") match {
case None => 0
case Some(title2) =>
- (StringUtilities.similarity(title1, title2) * MaxScore).toInt
+ (StringUtilities.similarity(title1.toLowerCase, title2.toLowerCase) * MaxScore).toInt
}
}
}