UrbComp2012 Paper21 Ying by aP68V1sd

VIEWS: 0 PAGES: 32

									       Urban Point-of-Interest Recommendation by
            Mining User Check-in Behaviors


Josh Jia-Ching Ying, Eric Hsueh-Chan Lu, Wen-Ning Kuo and Vincent S. Tseng

         Institute of Computer Science and Information Engineering
                       National Cheng Kung University
          No.1, University Road, Tainan City 701, Taiwan (R.O.C.)



                                                             Intelligent DataBase
                                                             System Lab, NCKU, Taiwan
                             Outline
     Introduction
        Background

        Motivation

        Challenges

     Proposed Method – UPOI-Mine

     Experimental Results

     Conclusions




                                       Intelligent DataBase
2                                      System Lab, NCKU, Taiwan
              Introduction – Background

     The markets of Location-Based Services (LBSs) in
      urban areas have grown rapidly.
     Effective and efficient urban POI recommendation
      techniques are desirable.
     Location Based Social Network (LBSN) data is widely
      used for building POI recommendation model.




                                               Intelligent DataBase
3                                              System Lab, NCKU, Taiwan
         Introduction – Background (cont.)

     heterogeneous data




                                      Intelligent DataBase
4                                     System Lab, NCKU, Taiwan
                  Introduction – Motivation
    We can not accurately catch users’ preference by
    analyzing his and his friend’s check-in actives




           ?                               ?


                                                       Intelligent DataBase
5                                                      System Lab, NCKU, Taiwan
                Introduction – Challenges
     How to understand user preference from LBSN data?
        How to extract useful features from heterogeneous data?




     How to precisely estimate the relevance between a user-
      POI pair based on the extracted features?
        How to integrate heterogeneous information?




                                                         Intelligent DataBase
6                                                        System Lab, NCKU, Taiwan
               Proposed Method – UPOI-Mine
Offline :UPOI-Mine


      Individual      POI popularity   Social Factor
    Preference (IP)       (PP)             (SF)

       Location         Check-in          Social
        Types            Data             Links


                      LBSN Dataset


Online: Recommender



                                                       Intelligent DataBase
7                                                      System Lab, NCKU, Taiwan
                        Feature Extraction
Offline :UPOI-Mine


      Individual      POI popularity   Social Factor
    Preference (IP)       (PP)             (SF)

       Location         Check-in          Social
        Types            Data             Links


                      LBSN Dataset


Online: Recommender



                                                       Intelligent DataBase
8                                                      System Lab, NCKU, Taiwan
                           Social Factor (SF)
    Weighted summation:
                                 F
                                                   Weight
      SF (useri ,POI j )  [ Interestk,j  Relation i,k ]
                                k 1

       Relation i ,k  w  CheckSimi ,k  (1  w)  DisSimi ,k

                           checkink , j
       Interestk , j    |S |

                          checkin
                         s 1
                                       k ,s

       F: friends of user i
       S: the set of POIs
       U: the set of user i’s friends
       Check-in k,* = check-ins of user k at POI*
                                                                 Intelligent DataBase
9                                                                System Lab, NCKU, Taiwan
                Social Factor – Relation

      Check-in Similarity (CheckSim)
         based on their check-in log
      Relative Distance Similarity (DisSim)
         based on their geographic distance




                                               Intelligent DataBase
10                                             System Lab, NCKU, Taiwan
                                  Relation – CheckSim

                                (1 0)  (0 10 )  (2  0)  (5 1)  (0  0)
     CheckSimi,j                                                                        0.0908
                            1  0  2  5  0  0  10  0  1  0
                            2       2    2    2       2    2       2   2    2       2


         Friend Indicator
              i      j          k       …         POI ID       A       B        C       D        E
                                                  user i       1       0        2       5        0
     i        0      1          0       …
                                                  user j       0       10       0       1        0
     j        1      0          1       …         user k       1       1        0       0        0
     k        0      1          0       ..        user l       1       1        1       1        1
     …        …      …          …       …         …            …       …        …       …        …




                                                                                 Intelligent DataBase
11                                                                               System Lab, NCKU, Taiwan
                                Relation – DisSim
          Distance  dissimilarity


          Maxi=1000
                          Distance                        Friend Indicator

             i        j          k       …                i       j          k       …
     i       0        100        10      …         i      0       1          0       …
     j       100      0          50      …         j      1       0          1       …
     k       10       50         0       ..        k      0       1          0       ..
     …       …        …          …       …         …      …       …          …       …
                                             100
                             DisSimi,j  1        0.9
                                             1000
                                                                      Intelligent DataBase
12                                                                    System Lab, NCKU, Taiwan
                              Social Factor – Example
  w  0.1
                                     10
  Social Factor from User B              [0.1 0.5  (1  0.1)  0.03]  0.0077
   Social Factor from User C  ..... 100
 Social Factor from User D  .....                          Relation:
   Social Factor from User E  .....
--------------------------------                             CheckSim(A, B) = 0.5
  social factorof user A to POI k       User A               DisSim(A, B) = 0.03

                                    ?
            POIk



                                                        User B
                                                        #Check-ins at POIK : 10
                                                        #Total Check-ins : 100
                                                                               10
                                                        Interest(B, POIK) =
                                                                               100
                                                                  Intelligent DataBase
 13                                                               System Lab, NCKU, Taiwan
                                   Individual Preference (IP)
                                                                                            highlight
                                                                 category
     • Individual Preference(IP)
         • HPrefi,h
         • CPrefi,c

      IP(user,POI j ) 
            i

                                                                                               
                                                                                 HCounth , j 
        C Pr ef i,c  I ctg ( c ) (POI j )   (1   )    H Pr ef i,h                  
        cC                                                 hH               H HCountg , j 
                                                                               g             
     , whereI(s,c) is an indicator function defined as

                              1    if POI j  ctg (c)
       I ctg ( c ) (POI j )  
                              0    otherwise
                                                                                 Intelligent DataBase
14                                                                               System Lab, NCKU, Taiwan
             Individual Preference – HPref & CPref
          POI                   A(c1)    B(c2)          C(c2)        D(c3)         Total
          Highlight             h1,h2    h1,h2          h2           h3
          Check-in count        5        1              2            2             10


 User      c1     c2       c3       c4                h1     h2              h3    h4        h5     Total
 User1     A      B,C      D             User1        A,B    A,B,C           D
                                         Total        5+1    5+1+2           2     0         0      16

           5 1 2 2                                   5 1 5  1  2               2 proportion of
                                                                                     check-ins of
          10 10 10                                    16      16                  16 the label
     Category           CPrefi,c                 Highlight        HPrefi,h

     C1                 0.5                      H1               0.375
     C2                 0.3                      H2               0.5
     C3                 0.2                      H3               0.125
                                                                                        Intelligent DataBase
15                                                                                      System Lab, NCKU, Taiwan
                 Individual Preference – Example
       There is only one category for one POI.        User A’s pref table
       There are many highlights for one POI.       Category           CPref
                                                     Seafood            0.5
                               Counts of highlight   Hotdog &           0.1
                                                     Sausages
                         POI
Category: Hotdog & Sausages                          Fast food          0.1
Highlight: Coffee(12), Cheese(88)                    Steak              0.3

                                                     Highlight          HPref
                                                     Coffee             0.5
                                                     Sightseeing        0.1
                                                     Ice Cream          0.1
                                                     Cheese             0.3

                                                                   Intelligent DataBase
 16                                                                System Lab, NCKU, Taiwan
           Individual Preference – Example (cont.)

                                                      User A’s pref table
                                                    Category           CPref
                        POI A                       Seafood            0.5
Category: Hotdog & Sausages                         Hotdog &           0.1
Highlight: Coffee(12), Cheese(88)                   Sausages
                                                    Fast food          0.1
         0.2                    CPref
                                                    Steak              0.3
       IP( UserA, POI j )  0.2  0.1 
                                                    Highlight          HPref
                          12             88 
       (1  0.2)  (0.5      )  (0.1     )     Coffee             0.5
                          100           100       Sightseeing        0.1
        0.168                              HPref
                                                    Ice Cream          0.1
                                                    Cheese             0.3
                                                                  Intelligent DataBase
 17                                                               System Lab, NCKU, Taiwan
                          POI Popularity (PP)
      POI Popularity
         Relative Popularity of POI

             Normalized based on category

                  checkinsj
       RPj 
                  checkins
               POIk CS
                              k


                                                                ith
       , whereCS is the set of POIs which in the same category w POI j.




                                                        Intelligent DataBase
18                                                      System Lab, NCKU, Taiwan
               POI Popularity – Example

                                        Hot Dogs    Check-in count
                Frank
     Category: Hot Dogs                 Frank       4,032
                                        KKK         25
                                        ……          …
                                        total       100,000




                                4,032
                   RPFrank              0.04032
                               100 ,000

                                                    Intelligent DataBase
19                                                  System Lab, NCKU, Taiwan
                       Relevance Estimation
 Offline :UPOI-Mine


       Individual      POI popularity   Social Factor
     Preference (IP)       (PP)             (SF)

        Location         Check-in          Social
         Types            Data             Links


                       LBSN Dataset


 Online: Recommender



                                                        Intelligent DataBase
20                                                      System Lab, NCKU, Taiwan
                 Relevance Estimation – Example
     To estimate the relevance of each pair of user to POI, Target
     we use these feature to learn a Regression-Tree Model.

       User ID      POI ID    SF       PP            IP             Relevance

       1            A         0.2      0.1           0.001          3
       1            B         0.05     0.2           0.1            5
       1            C         0.004    0.1           0.9            1
       …            …         …        …             …              …
       N            D         0.5      0.15          0.06           2




                                      Regression-Tree Model



                                                              Intelligent DataBase
21                                                            System Lab, NCKU, Taiwan
       Relevance Estimation – Regression-Tree
                      Model
      Regression-Tree Model has shown excellent performance
       for numerical value prediction
         • Demographic Prediction
         • Bio Life Cycle Analysis
         • Prediction of Geographical Natural
      Learning Steps:
         1. Building the initial tree
         2. Linear regression model for each leaf node
         3. Pruning the tree




                                                          Intelligent DataBase
22                                                        System Lab, NCKU, Taiwan
                            Recommender
 Offline :UPOI-Mine


       Individual      POI popularity   Social Factor
     Preference (IP)       (PP)             (SF)

        Location         Check-in          Social
         Types            Data             Links


                       LBSN Dataset


 Online: Recommender



                                                        Intelligent DataBase
23                                                      System Lab, NCKU, Taiwan
               Experimental Evaluation

      Experimental dataset – Gowalla Dataset
         Near or within New York City
         1,964,919 POIs
         18,159 people
         5,341,191 Check-ins
         392,246 Friendship Links




                                                Intelligent DataBase
24                                              System Lab, NCKU, Taiwan
                    Experimental Evaluation
      Experimental measurements
         Normalized Discounted Cumulative Gain (NDCG)

                     G[i ],                if i  1
                      DCG[i  1]  G[i ],   if i  b
                     
           DCG[i ]                                                       DCG [ p]
                                                            NDCG @ p 
                      DCG[i  1]  G[i ] , if i  b                      IDCG [ p]
                     
                                   log b i

              To measure ranking performance of relevance score of top k POIs in
               recommendation list


         Mean Absolute Error (MAE)
                     1 n
              MAE   f i  y i
                     n i 1
              To measure error of relevance score of all POIs
                                                                   Intelligent DataBase
25                                                                 System Lab, NCKU, Taiwan
              Experimental Evaluation (cont.)
                                             avg = 200
      Ground Truth
            x  avg                       POI ID        Check-in            Relevance
         3
         max -avg  2, if x  avg
                                          A             50                  1
            x-avg
        3 -          2, if x  avg       B             50                  1
         min -avg
        
                                           C             500                 5
                                           D             200                 3

      Baseline
         Trust Walker
              M. Jamali, M. Ester. TrustWalker: A Random Walk Model for Combining
               Trust-based and Item-based Recommendation. Proceedings of KDD,
               pages 397-406, Paris, 2009.

         Multi-Factor CF-based
              M. Ye, P. Yin, W.-C. Lee and Dik-Lun Lee. Exploiting Geographical
               Influence for Collaborative Point-of-Interest Recommendation.
               Proceedings of SIGIR, pages 1046-1054, Beijing, China, 2011.
                                                                       Intelligent DataBase
26                                                                         System Lab, NCKU, Taiwan
                     Comparison of Various Features
             The Individual Preference is more important than Social Factor for urban
              POI recommendation.

                                                          100%




                                                NDCG@10
      0.7                     0.65                         80%
              0.62    0.63                                 60%
MAE




                                      0.59
      0.6                                                  40%
                                                           20%
      0.5                                                   0%




                                                                         Intelligent DataBase
 27                                                                      System Lab, NCKU, Taiwan
      Comparison of Various Features (cont.)



          MAE                     100%
                                   80%




                        NDCG@10
0.7
                                   60%
0.6                                40%
                                   20%
0.5                                 0%




                                         Intelligent DataBase
28                                       System Lab, NCKU, Taiwan
                        Comparison with Existing
                            Recommenders

                                            NDCG@10

                                    TrustWalker

                          Multi-Factor CF-based

     Multi-Factor CF-based(geographic influence)

 Multi-Factor CF-based( user prefrence influence)

         Multi-Factor CF-based( social influence)

                              Our approach (PP)

                              Our approach (SF)

                               Our approach (IP)

                              Our approach (All)

                                                    0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

                                                                           Intelligent DataBase
29                                                                         System Lab, NCKU, Taiwan
                        Comparison with Existing
                         Recommenders (cont.)


                                              MAE
                                    TrustWalker
                          Multi-Factor CF-based
     Multi-Factor CF-based(geographic influence)
 Multi-Factor CF-based( user prefrence influence)
         Multi-Factor CF-based( social influence)
                              Our approach (PP)
                              Our approach (SF)
                               Our approach (IP)
                              Our approach (All)
                                                    0.00   0.50   1.00   1.50         2.00        2.50




                                                                                Intelligent DataBase
30                                                                              System Lab, NCKU, Taiwan
                         Conclusions
 We proposed a novel urban POIs recommendation which is
  called UPOI-Mine by mining users’ preferences.
 we propose three kinds of useful features
   Social Factor
   Individual Preference
   POI Popularity
 Through a series of experiments by the real dataset Gowalla
   We have validated our proposed UPOI-Mine and shown that
    UPOI-Mine has excellent performance under various
    conditions.
   The Individual Preference is more important than Social
    Factor for urban POI recommendation.
                                                         Intelligent DataBase
                                                         System Lab, NCKU, Taiwan
Thank you for your attentions


          Question?




                                Intelligent DataBase
                                System Lab, NCKU, Taiwan

								
To top