diff --git a/changelog.md b/changelog.md
--- a/changelog.md
+++ b/changelog.md
@@ -1,3 +1,8 @@
+= 0.6.0.0 =
+
+ - Switched to using Hashmap for the DefaultMap implementation;
+   this *may* break compatibility with old binary stores.
+
 = 0.5.2.1 =
 
  - Removed an orphan instance for Text.
diff --git a/chatter.cabal b/chatter.cabal
--- a/chatter.cabal
+++ b/chatter.cabal
@@ -1,5 +1,5 @@
 name:                chatter
-version:             0.5.2.1
+version:             0.6.0.0
 synopsis:            A library of simple NLP algorithms.
 description:         chatter is a collection of simple Natural Language
                      Processing algorithms.
@@ -78,6 +78,7 @@
                      random-shuffle >= 0.0.4,
                      MonadRandom    >= 0.1.2,
                      cereal         >= 0.4.0.1,
+                     cereal-text    >= 0.1 && < 0.2,
                      fullstop       >= 0.1.3.1,
                      bytestring     >= 0.10.0.0,
                      directory,
@@ -92,7 +93,9 @@
                      regex-tdfa-text,
                      array,
                      QuickCheck,
-                     quickcheck-instances
+                     quickcheck-instances,
+                     unordered-containers,
+                     hashable
 
 
 
diff --git a/src/Data/DefaultMap.hs b/src/Data/DefaultMap.hs
--- a/src/Data/DefaultMap.hs
+++ b/src/Data/DefaultMap.hs
@@ -2,10 +2,12 @@
 module Data.DefaultMap
 where
 
+import Control.Applicative ((<$>), (<*>))
 import Test.QuickCheck (Arbitrary(..))
 import Control.DeepSeq (NFData)
-import Data.Map (Map)
-import qualified Data.Map as Map
+import Data.Hashable
+import Data.HashMap.Strict (HashMap)
+import qualified Data.HashMap.Strict as Map
 import Data.Serialize
 import GHC.Generics
 
@@ -13,12 +15,15 @@
 -- for a key that does not exist.
 data DefaultMap k v =
    DefMap { defDefault :: v
-          , defMap :: Map k v
-          } deriving (Read, Show, Eq, Ord, Generic)
+          , defMap :: HashMap k v
+          } deriving (Read, Show, Eq, Generic)
 
-instance (Ord k, Serialize k, Serialize v) => Serialize (DefaultMap k v)
-instance (NFData k, NFData v, Ord k) => NFData (DefaultMap k v)
+instance (Eq k, Hashable k, Serialize k, Serialize v) => Serialize (DefaultMap k v) where
+  put (DefMap d theMap) = put d >> put (Map.toList theMap)
+  get = DefMap <$> get <*> (Map.fromList <$> get)
 
+instance (NFData k, NFData v, Hashable k) => NFData (DefaultMap k v)
+
 -- | Create an empty `DefaultMap`
 empty :: v -> DefaultMap k v
 empty def = DefMap { defDefault = def
@@ -26,11 +31,11 @@
 
 -- | Query the map for a value.  Returns the default if the key is not
 -- found.
-lookup :: Ord k => k -> DefaultMap k v ->  v
-lookup k m = Map.findWithDefault (defDefault m) k (defMap m)
+lookup :: (Eq k, Hashable k) => k -> DefaultMap k v ->  v
+lookup k m = Map.lookupDefault (defDefault m) k (defMap m)
 
 -- | Create a `DefaultMap` from a default value and a list.
-fromList :: Ord k => v -> [(k, v)] -> DefaultMap k v
+fromList :: (Eq k, Hashable k) => v -> [(k, v)] -> DefaultMap k v
 fromList def entries = DefMap { defDefault = def
                               , defMap = Map.fromList entries }
 
@@ -44,9 +49,9 @@
 -- over the default value -- this fold behaves in the same way as a
 -- standard `Data.Map.foldl`
 foldl :: (a -> b -> a) -> a -> DefaultMap k b -> a
-foldl fn acc m = Map.foldl fn acc (defMap m)
+foldl fn acc m = Map.foldl' fn acc (defMap m)
 
-instance (Arbitrary k, Arbitrary v, Ord k) => Arbitrary (DefaultMap k v) where
+instance (Arbitrary k, Arbitrary v, Hashable k, Eq k) => Arbitrary (DefaultMap k v) where
   arbitrary = do
       def <- arbitrary
       entries <- arbitrary
diff --git a/src/NLP/Similarity/VectorSim.hs b/src/NLP/Similarity/VectorSim.hs
--- a/src/NLP/Similarity/VectorSim.hs
+++ b/src/NLP/Similarity/VectorSim.hs
@@ -20,7 +20,8 @@
 
 
 -- | Invokes similarity on full strings, using `T.words` for
--- tokenization, and no stemming.
+-- tokenization, and no stemming. The return value will be in the
+-- range [0, 1]
 --
 -- There *must* be at least one document in the corpus.
 sim :: Corpus -> Text -> Text -> Double
@@ -35,6 +36,8 @@
 -- implementation.  If you need to run similarity against any single
 -- document more than once, then you should create `TermVector`s for
 -- each of your documents and use `tvSim` instead of `similarity`.
+-- 
+-- The return value will be in the range [0, 1].
 --
 -- There *must* be at least one document in the corpus.
 similarity :: Corpus -> [Text] -> [Text] -> Double
@@ -46,7 +49,8 @@
 -- | Determine how similar two documents are.
 --
 -- Calculates the similarity between two documents, represented as
--- `TermVectors`
+-- `TermVectors`, returning a double in the range [0, 1] where 1 represents
+-- "most similar".
 tvSim :: TermVector -> TermVector -> Double
 tvSim doc1 doc2 = let
   theCos = cosVec doc1 doc2
diff --git a/src/NLP/Types/Tags.hs b/src/NLP/Types/Tags.hs
--- a/src/NLP/Types/Tags.hs
+++ b/src/NLP/Types/Tags.hs
@@ -1,13 +1,12 @@
 {-# LANGUAGE DeriveGeneric #-}
 {-# LANGUAGE OverloadedStrings #-}
-{-# OPTIONS_GHC -fno-warn-orphans #-}
 module NLP.Types.Tags
 where
 
-import Data.Serialize (Serialize, get, put)
+import Data.Serialize (Serialize)
+import Data.Serialize.Text ()
 import Data.Text (Text)
 import qualified Data.Text as T
-import Data.Text.Encoding (encodeUtf8, decodeUtf8)
 import GHC.Generics
 import Text.Read (readEither)
 
@@ -97,8 +96,3 @@
   arbitrary = do
     NonEmpty str <- arbitrary
     return $ RawTag $ T.pack str
-
-instance Serialize Text where
-  put txt = put $ encodeUtf8 txt
-  get     = fmap decodeUtf8 get
-