diff --git a/Scrapy/pipelines.py b/Scrapy/pipelines.py
index 3345787..3194d7e 100644
--- a/Scrapy/pipelines.py
+++ b/Scrapy/pipelines.py
@@ -2,7 +2,24 @@
 #
 # Don't forget to add your pipeline to the ITEM_PIPELINES setting
 # See: http://doc.scrapy.org/en/latest/topics/item-pipeline.html
+from scrapy.exceptions import DropItem
+
 
 class FourmiPipeline(object):
+
+    def __init__(self):
+        self.known_values = set()
+
     def process_item(self, item, spider):
-        return item
+        """
+        Processing the items so exact doubles are dropped
+        :param item: The incoming item
+        :param spider: The spider which scraped the spider
+        :return: :raise DropItem: Returns the item if unique or drops them if it's already known
+        """
+        value = item['attribute'], item['value']
+        if value in self.known_values:
+            raise DropItem("Duplicate item found: %s" % item)
+        else:
+            self.known_values.add(value)
+            return item
diff --git a/Scrapy/settings.py b/Scrapy/settings.py
index e43aa2b..fd379a9 100644
--- a/Scrapy/settings.py
+++ b/Scrapy/settings.py
@@ -10,6 +10,9 @@ BOT_NAME = 'Fourmi'
 
 SPIDER_MODULES = ['Scrapy.spiders']
 NEWSPIDER_MODULE = 'Scrapy.spiders'
+ITEM_PIPELINES = {
+    'Scrapy.pipelines.FourmiPipeline': 100
+}
 
 # Crawl responsibly by identifying yourself (and your website) on the user-agent
 #USER_AGENT = 'Fourmi (+http://www.yourdomain.com)'