From b4eac17049e19d33b1a55664a7258c0f62f8a8c7 Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Mon, 27 Aug 2018 16:40:15 -0700 Subject: make similarity score case-insensitive --- scalding/src/main/scala/sandcrawler/Scorable.scala | 2 +- scalding/src/test/scala/sandcrawler/ScorableTest.scala | 8 ++++++++ 2 files changed, 9 insertions(+), 1 deletion(-) (limited to 'scalding') 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 } } } diff --git a/scalding/src/test/scala/sandcrawler/ScorableTest.scala b/scalding/src/test/scala/sandcrawler/ScorableTest.scala index f63bef8..2094543 100644 --- a/scalding/src/test/scala/sandcrawler/ScorableTest.scala +++ b/scalding/src/test/scala/sandcrawler/ScorableTest.scala @@ -70,4 +70,12 @@ class ScorableTest extends FlatSpec with Matchers { new ReduceFeatures(JsonString), new ReduceFeatures(JsonString)) score shouldBe Scorable.MaxScore } + + "computeOutput()" should "be case-insensitive" in { + val left = JsonString.replace("<>", "A TITLE UPPER CASE") + val right = JsonString.replace("<<TITLE>>", "a title upper case") + val score = Scorable.computeSimilarity( + new ReduceFeatures(left), new ReduceFeatures(right)) + score shouldBe Scorable.MaxScore + } } -- cgit v1.2.3