Difference between revisions of "Collaborative filtering"

From HaskellWiki
Jump to navigation Jump to search
m
m (CollaborativeFiltering moved to Collaborative filtering)
 
(5 intermediate revisions by 2 users not shown)
Line 1: Line 1:
This page was added to discuss different versions of the code for collaborative filtering at [http://www.serpentine.com/blog/2007/08/27/weighted-slope-one-in-haskell-collaborative-filtering-in-29-lines-of-code/#comment-75580 Bryan's blog].
+
[[Category:Code]] [[Category:Idioms]]This page was added to discuss different versions of the code for collaborative filtering at [http://www.serpentine.com/blog/2007/08/27/weighted-slope-one-in-haskell-collaborative-filtering-in-29-lines-of-code/#comment-75580 Bryan's blog].
   
 
== Chris' version ==
 
== Chris' version ==
Line 5: Line 5:
 
I renamed the variables and then reorganized the code a bit.
 
I renamed the variables and then reorganized the code a bit.
   
  +
The(update,SlopeOne,predict) definitions can be compared with the newer (update2,SlopeOne',predict') definitions.
The predict' function replaces predict.
 
The update2
 
   
 
<haskell>
 
<haskell>
Line 21: Line 20:
 
type Rating item = M.Map item RatingValue
 
type Rating item = M.Map item RatingValue
   
-- The SlopeOne matrix is indexed by pairs of items and is implmeneted
+
-- The SlopeOne matrix is indexed by pairs of items and is implemented
-- as a sparse map of maps. If the item type is an instance of Show
+
-- as a sparse map of maps.
-- then so is the (SlopeOne item) type.
 
 
newtype SlopeOne item = SlopeOne (M.Map item (M.Map item (Count,RatingValue)))
 
newtype SlopeOne item = SlopeOne (M.Map item (M.Map item (Count,RatingValue)))
 
deriving (Show)
 
deriving (Show)
   
-- The SlopeOne' matrix is an unormalized version of SlopeOne
+
-- The SlopeOne' matrix is an un-normalized version of SlopeOne
 
newtype SlopeOne' item = SlopeOne' (M.Map item (M.Map item (Count,RatingValue)))
 
newtype SlopeOne' item = SlopeOne' (M.Map item (M.Map item (Count,RatingValue)))
 
deriving (Show)
 
deriving (Show)
Line 57: Line 55:
 
un_norm (count,rating) = (count, rating * fromIntegral count)
 
un_norm (count,rating) = (count, rating * fromIntegral count)
 
matrixIn = M.map (M.map un_norm) matrixInNormed
 
matrixIn = M.map (M.map un_norm) matrixInNormed
  +
   
 
-- This version of update2 makes an unnormalize slopeOne' from each
 
-- This version of update2 makes an unnormalize slopeOne' from each
 
-- Rating and combines them using Map.union* operations and addT.
 
-- Rating and combines them using Map.union* operations and addT.
 
update2 :: Ord item => SlopeOne' item -> [Rating item] -> SlopeOne' item
 
update2 :: Ord item => SlopeOne' item -> [Rating item] -> SlopeOne' item
update2 s [] = s
+
update2 s@(SlopeOne' matrixIn) usersRatingsIn | null usersRatings = s
  +
| otherwise =
update2 (SlopeOne' matrixIn) usersRatings =
 
 
SlopeOne' . M.unionsWith (M.unionWith addT) . (matrixIn:) . map fromRating $ usersRatings
 
SlopeOne' . M.unionsWith (M.unionWith addT) . (matrixIn:) . map fromRating $ usersRatings
  +
where usersRatings = filter ((1<) . M.size) usersRatingsIn
where fromRating userRating = M.mapWithKey expand1 userRating
+
fromRating userRating = M.mapWithKey expand1 userRating
 
where expand1 item1 rating1 = M.mapMaybeWithKey expand2 userRating
 
where expand1 item1 rating1 = M.mapMaybeWithKey expand2 userRating
 
where expand2 item2 rating2 | item1 == item2 = Nothing
 
where expand2 item2 rating2 | item1 == item2 = Nothing
 
| otherwise = Just (1,rating1 - rating2)
 
| otherwise = Just (1,rating1 - rating2)
+
 
predict :: Ord a => SlopeOne a -> Rating a -> Rating a
 
predict :: Ord a => SlopeOne a -> Rating a -> Rating a
 
predict (SlopeOne matrixIn) userRatings =
 
predict (SlopeOne matrixIn) userRatings =
Line 89: Line 89:
 
-- unnormalized SlopeOne' but this is a small detail
 
-- unnormalized SlopeOne' but this is a small detail
 
predict' :: Ord a => SlopeOne' a -> Rating a -> Rating a
 
predict' :: Ord a => SlopeOne' a -> Rating a -> Rating a
predict' (SlopeOne' matrixIn) userRatings = M.mapMaybeWithKey calcItem matrixIn
+
predict' (SlopeOne' matrixIn) userRatings =
where calcItem item1 innerMap | M.member item1 userRatings = Nothing
+
M.mapMaybe calcItem (M.difference matrixIn userRatings)
| M.null combined = Nothing
+
where calcItem innerMap | M.null combined = Nothing
| norm_rating <= 0 = Nothing
+
| norm_rating <= 0 = Nothing
| otherwise = Just norm_rating
+
| otherwise = Just norm_rating
 
where combined = M.intersectionWith weight innerMap userRatings
 
where combined = M.intersectionWith weight innerMap userRatings
 
(total_count,total_rating) = foldl1' addT (M.elems combined)
 
(total_count,total_rating) = foldl1' addT (M.elems combined)
Line 108: Line 108:
 
]
 
]
   
userInfo = M.fromList [("squid", 0.4)]
+
userInfo = M.fromList [("squid", 0.4),("cuttlefish",0.9),("dolphin",1.0)]
   
 
predictions = predict (update empty userData) userInfo
 
predictions = predict (update empty userData) userInfo
   
 
predictions' = predict' (update2 empty' userData) userInfo
 
predictions' = predict' (update2 empty' userData) userInfo
  +
</haskell>
  +
  +
== More optimized storage ==
  +
  +
The changes to SlopeOne/update/predict below use a different internal data structure for storing the sparse matrix of SlopeOne. Instead of a Map of Map design it uses a Map of List design and keeps the List in distinct ascending form. The list values are a strict (data Tup) type which should help save space compared to the previous inner Map design and yet efficiently provide all the operations needed by update and predict.
  +
  +
Much of the logic of prediction is in the computeRating helper function.
  +
  +
<haskell>
  +
-- The SlopeOne matrix is indexed by pairs of items and is implemented
  +
-- as a sparse map of distinct ascending lists. The 'update' and
  +
-- 'predict' functions do not need the inner type to actually be a
  +
-- map, so the list saves space and complexity.
  +
newtype SlopeOne item = SlopeOne (M.Map item [Tup item])
  +
deriving (Show)
  +
  +
-- Strict triple tuple type for SlopeOne internals
  +
data Tup item = Tup { itemT :: !item, countT :: !Count, ratingT :: !RatingValue }
  +
deriving (Show)
  +
  +
empty :: SlopeOne item
  +
empty = SlopeOne M.empty
  +
  +
update :: Ord item => SlopeOne item -> [Rating item] -> SlopeOne item
  +
update s@(SlopeOne matrixIn) usersRatingsIn | null usersRatings = s
  +
| otherwise =
  +
SlopeOne . M.unionsWith mergeAdd . (matrixIn:) . map fromRating $ usersRatings
  +
where usersRatings = filter ((1<) . M.size) usersRatingsIn
  +
-- fromRating converts a Rating into a Map of Lists, a singleton SlopeOne.
  +
fromRating userRatings = M.mapWithKey expand userRatings
  +
where expand item1 rating1 = map makeTup . M.toAscList . M.delete item1 $ userRatings
  +
where makeTup (item2,rating2) = Tup item2 1 (rating1-rating2)
  +
  +
-- 'mergeAdd' is a helper for 'update'.
  +
-- Optimized traversal of distinct ascending lists to perform additive merge.
  +
mergeAdd :: Ord item => [Tup item] -> [Tup item] -> [Tup item]
  +
mergeAdd !xa@(x:xs) !ya@(y:ys) =
  +
case compare (itemT x) (itemT y) of
  +
LT -> x : mergeAdd xs ya
  +
GT -> y : mergeAdd xa ys
  +
EQ -> Tup (itemT x) (countT x + countT y) (ratingT x + ratingT y) : mergeAdd xs ys
  +
mergeAdd xs [] = xs
  +
mergeAdd [] ys = ys
  +
  +
-- The output Rating has no items in common with the input Rating and
  +
-- only includes positively weighted ratings.
  +
predict :: Ord item => SlopeOne item -> Rating item -> Rating item
 
predict (SlopeOne matrixIn) userRatings =
  +
M.mapMaybe (computeRating ratingList) (M.difference matrixIn userRatings)
  +
where ratingList = M.toAscList userRatings
  +
  +
-- 'computeRating' is a helper for 'predict'.
  +
-- Optimized traversal of distinct ascending lists to compute positive weighted rating.
  +
computeRating :: (Ord item) => [(item,RatingValue)] -> [Tup item] -> Maybe RatingValue
  +
computeRating !xa@(x:xs) !ya@(y:ys) =
  +
case compare (fst x) (itemT y) of
  +
LT -> computeRating xs ya
  +
GT -> computeRating xa ys
  +
EQ -> helper (countT y) (ratingT y + fromIntegral (countT y) * snd x) xs ys
  +
where
  +
helper :: (Ord item) => Count -> RatingValue -> [(item,RatingValue)] -> [Tup item] -> Maybe RatingValue
  +
helper !count !rating !xa@(x:xs) !ya@(y:ys) =
  +
case compare (fst x) (itemT y) of
  +
LT -> helper count rating xs ya
  +
GT -> helper count rating xa ys
  +
EQ -> helper (count + countT y) (rating + ratingT y + fromIntegral (countT y) * (snd x)) xs ys
  +
helper !count !rating _ _ | rating > 0 = Just (rating / fromIntegral count)
  +
| otherwise = Nothing
  +
computeRating _ _ = Nothing
 
</haskell>
 
</haskell>

Latest revision as of 15:39, 20 April 2008

This page was added to discuss different versions of the code for collaborative filtering at Bryan's blog.

Chris' version

I renamed the variables and then reorganized the code a bit.

The(update,SlopeOne,predict) definitions can be compared with the newer (update2,SlopeOne',predict') definitions.

module WeightedSlopeOne (Rating, SlopeOne, empty, predict, update) where

import Data.List (foldl',foldl1')
import qualified Data.Map as M

-- The item type is a polymorphic parameter.  Since it goes into a Map
-- it must be able to be compared, so item must be an instance of Ord.
type Count = Int
type RatingValue = Double
-- The Rating is the known (item,Rating) information for a particular "user"
type Rating item = M.Map item RatingValue

-- The SlopeOne matrix is indexed by pairs of items and is implemented
-- as a sparse map of maps.
newtype SlopeOne item = SlopeOne (M.Map item (M.Map item (Count,RatingValue)))
  deriving (Show)

-- The SlopeOne' matrix is an un-normalized version of SlopeOne
newtype SlopeOne' item = SlopeOne' (M.Map item (M.Map item (Count,RatingValue)))
  deriving (Show)

empty = SlopeOne M.empty
empty' = SlopeOne' M.empty

-- This performs a strict addition on pairs made of two nuumeric types
addT (a,b) (c,d) = let (l,r) = (a+c, b+d) in l `seq` r `seq` (l, r)


-- There is never an entry for the "diagonal" elements with equal
-- items in the pair: (foo,foo) is never in the SlopeOne.
update :: Ord item => SlopeOne item -> [Rating item] -> SlopeOne item
update (SlopeOne matrixInNormed) usersRatings =
    SlopeOne . M.map (M.map norm) . foldl' update' matrixIn $ usersRatings
  where update' oldMatrix userRatings =
          foldl' (\oldMatrix (itemPair, rating) -> insert oldMatrix itemPair rating)
                 oldMatrix itemCombos
          where itemCombos = [ ((item1, item2), (1, rating1 - rating2)) 
                             | (item1, rating1) <- ratings
                             , (item2, rating2) <- ratings
                             , item1 /= item2]
                ratings = M.toList userRatings
        insert outerMap (item1, item2) newRating = M.insertWith' outer item1 newOuterEntry outerMap
          where newOuterEntry = M.singleton item2 newRating
                outer _ innerMap = M.insertWith' addT item2 newRating innerMap
        norm (count,total_rating) = (count, total_rating / fromIntegral count)
        un_norm (count,rating) = (count, rating * fromIntegral count)
        matrixIn = M.map (M.map un_norm) matrixInNormed


-- This version of update2 makes an unnormalize slopeOne' from each
-- Rating and combines them using Map.union* operations and addT.
update2 :: Ord item => SlopeOne' item -> [Rating item] -> SlopeOne' item
update2 s@(SlopeOne' matrixIn) usersRatingsIn | null usersRatings = s
                                              | otherwise =
    SlopeOne' . M.unionsWith (M.unionWith addT) . (matrixIn:) . map fromRating $ usersRatings
  where usersRatings = filter ((1<) . M.size) usersRatingsIn
        fromRating userRating = M.mapWithKey expand1 userRating
          where expand1 item1 rating1 = M.mapMaybeWithKey expand2 userRating
                  where expand2 item2 rating2 | item1 == item2 = Nothing
                                              | otherwise = Just (1,rating1 - rating2)
        
predict :: Ord a => SlopeOne a -> Rating a -> Rating a
predict (SlopeOne matrixIn) userRatings =
  let freqM = foldl' insert M.empty
                     [ (item1,found_rating,user_rating)
                     | (item1,innerMap) <- M.assocs matrixIn
                     , M.notMember item1 userRatings
                     , (user_item, user_rating) <- M.toList userRatings
                     , item1 /= user_item
                     , found_rating <- M.lookup user_item innerMap
                     ]
      insert oldM (item1,found_rating,user_rating) =
        let (count,norm_rating) = found_rating
            total_rating = fromIntegral count * (norm_rating + user_rating)
        in M.insertWith' addT item1 (count,total_rating) oldM
      normM = M.map (\(count, total_rating) -> total_rating / fromIntegral count) freqM
  in M.filter (\norm_rating -> norm_rating > 0) normM

-- This is a modified version of predict.  It also expect the
-- unnormalized SlopeOne' but this is a small detail
predict' :: Ord a => SlopeOne' a -> Rating a -> Rating a
predict' (SlopeOne' matrixIn) userRatings =
    M.mapMaybe calcItem (M.difference matrixIn userRatings)
  where calcItem innerMap | M.null combined = Nothing
                          | norm_rating <= 0 = Nothing
                          | otherwise = Just norm_rating
          where combined = M.intersectionWith weight innerMap userRatings
                (total_count,total_rating) = foldl1' addT (M.elems combined)
                norm_rating = total_rating / fromIntegral total_count
        weight (count,rating) user_rating =
          (count,rating + fromIntegral count *  user_rating)

userData :: [Rating String]
userData = map M.fromList [
 [("squid", 1.0), ("cuttlefish", 0.5), ("octopus", 0.2)],
 [("squid", 1.0), ("octopus", 0.5), ("nautilus", 0.2)],
 [("squid", 0.2), ("octopus", 1.0), ("cuttlefish", 0.4), ("nautilus", 0.4)],
 [("cuttlefish", 0.9), ("octopus", 0.4), ("nautilus", 0.5)]
 ]

userInfo = M.fromList [("squid", 0.4),("cuttlefish",0.9),("dolphin",1.0)]

predictions = predict (update empty userData) userInfo

predictions' = predict' (update2 empty' userData) userInfo

More optimized storage

The changes to SlopeOne/update/predict below use a different internal data structure for storing the sparse matrix of SlopeOne. Instead of a Map of Map design it uses a Map of List design and keeps the List in distinct ascending form. The list values are a strict (data Tup) type which should help save space compared to the previous inner Map design and yet efficiently provide all the operations needed by update and predict.

Much of the logic of prediction is in the computeRating helper function.

-- The SlopeOne matrix is indexed by pairs of items and is implemented
-- as a sparse map of distinct ascending lists.  The 'update' and
-- 'predict' functions do not need the inner type to actually be a
-- map, so the list saves space and complexity.
newtype SlopeOne item = SlopeOne (M.Map item [Tup item])
  deriving (Show)

-- Strict triple tuple type for SlopeOne internals
data Tup item = Tup { itemT :: !item, countT :: !Count, ratingT :: !RatingValue }
  deriving (Show)

empty :: SlopeOne item
empty = SlopeOne M.empty

update :: Ord item => SlopeOne item -> [Rating item] -> SlopeOne item
update s@(SlopeOne matrixIn) usersRatingsIn | null usersRatings = s
                                            | otherwise =
    SlopeOne . M.unionsWith mergeAdd . (matrixIn:) . map fromRating $ usersRatings
  where usersRatings = filter ((1<) . M.size) usersRatingsIn
        -- fromRating converts a Rating into a Map of Lists, a singleton SlopeOne.
        fromRating userRatings = M.mapWithKey expand userRatings
          where expand item1 rating1 = map makeTup . M.toAscList . M.delete item1 $ userRatings
                  where makeTup (item2,rating2) = Tup item2 1 (rating1-rating2)

-- 'mergeAdd' is a helper for 'update'.
-- Optimized traversal of distinct ascending lists to perform additive merge.
mergeAdd :: Ord item => [Tup item] -> [Tup item] -> [Tup item]
mergeAdd !xa@(x:xs) !ya@(y:ys) =
  case compare (itemT x) (itemT y) of
    LT -> x : mergeAdd xs ya
    GT -> y : mergeAdd xa ys
    EQ -> Tup (itemT x) (countT x + countT y) (ratingT x + ratingT y) : mergeAdd xs ys
mergeAdd xs [] = xs
mergeAdd [] ys = ys

-- The output Rating has no items in common with the input Rating and
-- only includes positively weighted ratings.
predict :: Ord item => SlopeOne item -> Rating item -> Rating item
predict (SlopeOne matrixIn) userRatings =
    M.mapMaybe (computeRating ratingList) (M.difference matrixIn userRatings)
  where ratingList = M.toAscList userRatings

-- 'computeRating' is a helper for 'predict'.
-- Optimized traversal of distinct ascending lists to compute positive weighted rating.
computeRating :: (Ord item) => [(item,RatingValue)] -> [Tup item] -> Maybe RatingValue
computeRating !xa@(x:xs) !ya@(y:ys) =
  case compare (fst x) (itemT y) of
    LT -> computeRating xs ya
    GT -> computeRating xa ys
    EQ -> helper (countT y) (ratingT y + fromIntegral (countT y) * snd x) xs ys
 where
  helper :: (Ord item) => Count -> RatingValue -> [(item,RatingValue)] -> [Tup item] -> Maybe RatingValue
  helper !count !rating !xa@(x:xs) !ya@(y:ys) =
    case compare (fst x) (itemT y) of
      LT -> helper count rating xs ya
      GT -> helper count rating xa ys
      EQ -> helper (count + countT y) (rating + ratingT y + fromIntegral (countT y) * (snd x)) xs ys
  helper !count !rating _ _  | rating > 0 = Just (rating / fromIntegral count)
                             | otherwise = Nothing
computeRating _ _ = Nothing