Document (#40825)

Author

Bessa, A.
Santos, R.L.T.
Veloso, A.
Ziviani, N.

Title

Exploiting item co-utility to improve collaborative filtering recommendations

Source

Journal of the Association for Information Science and Technology. 68(2017) no.10, S.2380-2393

Year

2017

Abstract

In this article we study the extent to which the interplay between recommended items affect recommendation effectiveness. We introduce and formalize the concept of co-utility as the property that any pair of recommended items has of being useful to a user, and exploit it to improve collaborative filtering recommendations. We present different techniques to estimate co-utility probabilities, all of them independent of content information, and compare them with each other. We use these probabilities, as well as normalized predicted ratings, in an instance of an NP-hard problem termed the Max-Sum Dispersion Problem (MSDP). A solution to MSDP hence corresponds to a set of items for recommendation. We study one heuristic and one exact solution to MSDP and perform comparisons among them. We also contrast our solutions (the best heuristic to MSDP) to different baselines by comparing the ratings users give to different recommendations. We obtain expressive gains in the utility of recommendations and our solutions also recommend higher-rated items to the majority of users. Finally, we show that our co-utility solutions are scalable in practice and do not harm recommendations' diversity.

Content

Vgl.: http://onlinelibrary.wiley.com/doi/10.1002/asi.23853/full.

Similar documents (author)

1. Brandão, W.C.; Santos, R.L.T.; Ziviani, N.; Moura, E.S. de; Silva, A.S. da: Learning to expand queries using entities (2014) 3.22

   3.2158647 = sum of:
     3.2158647 = sum of:
       1.2647046 = weight(author_txt:santos in 1343) [ClassicSimilarity], result of:
         1.2647046 = score(doc=1343,freq=1.0), product of:
           0.599472 = queryWeight, product of:
             8.43879 = idf(docFreq=25, maxDocs=44218)
             0.071037665 = queryNorm
           2.1096976 = fieldWeight in 1343, product of:
             1.0 = tf(freq=1.0), with freq of:
               1.0 = termFreq=1.0
             8.43879 = idf(docFreq=25, maxDocs=44218)
             0.25 = fieldNorm(doc=1343)
       1.9511602 = weight(author_txt:ziviani in 1343) [ClassicSimilarity], result of:
         1.9511602 = score(doc=1343,freq=1.0), product of:
           0.8003957 = queryWeight, product of:
             1.1554947 = boost
             9.7509775 = idf(docFreq=6, maxDocs=44218)
             0.071037665 = queryNorm
           2.4377444 = fieldWeight in 1343, product of:
             1.0 = tf(freq=1.0), with freq of:
               1.0 = termFreq=1.0
             9.7509775 = idf(docFreq=6, maxDocs=44218)
             0.25 = fieldNorm(doc=1343)
   

2. Santos, E.M.O.: CD-ROM em bibliotecas e servicos de informacoa (1995) 1.58

   1.5808808 = sum of:
     1.5808808 = product of:
       3.1617615 = sum of:
         3.1617615 = weight(author_txt:santos in 1597) [ClassicSimilarity], result of:
           3.1617615 = score(doc=1597,freq=1.0), product of:
             0.599472 = queryWeight, product of:
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.071037665 = queryNorm
             5.274244 = fieldWeight in 1597, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.625 = fieldNorm(doc=1597)
       0.5 = coord(1/2)
   

3. Green, L. Santos => Santos Green, L.: 1.34

   1.3414218 = sum of:
     1.3414218 = product of:
       2.6828437 = sum of:
         2.6828437 = weight(author_txt:santos in 4512) [ClassicSimilarity], result of:
           2.6828437 = score(doc=4512,freq=2.0), product of:
             0.599472 = queryWeight, product of:
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.071037665 = queryNorm
             4.4753447 = fieldWeight in 4512, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.375 = fieldNorm(doc=4512)
       0.5 = coord(1/2)
   

4. Bisi, P.H. Santos => Santos Bisi, P.H.: 1.34

   1.3414218 = sum of:
     1.3414218 = product of:
       2.6828437 = sum of:
         2.6828437 = weight(author_txt:santos in 4802) [ClassicSimilarity], result of:
           2.6828437 = score(doc=4802,freq=2.0), product of:
             0.599472 = queryWeight, product of:
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.071037665 = queryNorm
             4.4753447 = fieldWeight in 4802, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.375 = fieldNorm(doc=4802)
       0.5 = coord(1/2)
   

5. Hornung, C.; Santos, A.: ¬A proposal for a reference model for cooperative hypermedia systems (1992) 1.26

   1.2647046 = sum of:
     1.2647046 = product of:
       2.5294092 = sum of:
         2.5294092 = weight(author_txt:santos in 2059) [ClassicSimilarity], result of:
           2.5294092 = score(doc=2059,freq=1.0), product of:
             0.599472 = queryWeight, product of:
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.071037665 = queryNorm
             4.219395 = fieldWeight in 2059, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               8.43879 = idf(docFreq=25, maxDocs=44218)
               0.5 = fieldNorm(doc=2059)
       0.5 = coord(1/2)
   

Similar documents (content)

1. Shani, G.; Rokach, L.; Shapira, B.; Hadash, S.; Tangi, M.: Investigating confidence displays for top-N recommendations (2013) 0.18

   0.17684692 = sum of:
     0.17684692 = product of:
       0.7368622 = sum of:
         0.034787875 = weight(abstract_txt:different in 1136) [ClassicSimilarity], result of:
           0.034787875 = score(doc=1136,freq=2.0), product of:
             0.08589926 = queryWeight, product of:
               1.4088932 = boost
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.01663328 = queryNorm
             0.40498453 = fieldWeight in 1136, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
         0.039540842 = weight(abstract_txt:them in 1136) [ClassicSimilarity], result of:
           0.039540842 = score(doc=1136,freq=1.0), product of:
             0.11787221 = queryWeight, product of:
               1.6503994 = boost
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.01663328 = queryNorm
             0.33545518 = fieldWeight in 1136, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
         0.10200134 = weight(abstract_txt:recommended in 1136) [ClassicSimilarity], result of:
           0.10200134 = score(doc=1136,freq=1.0), product of:
             0.19368128 = queryWeight, product of:
               1.7273549 = boost
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.01663328 = queryNorm
             0.5266453 = fieldWeight in 1136, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
         0.20230682 = weight(abstract_txt:recommendation in 1136) [ClassicSimilarity], result of:
           0.20230682 = score(doc=1136,freq=3.0), product of:
             0.2119908 = queryWeight, product of:
               1.8071585 = boost
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.01663328 = queryNorm
             0.9543189 = fieldWeight in 1136, product of:
               1.7320508 = tf(freq=3.0), with freq of:
                 3.0 = termFreq=3.0
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
         0.20027743 = weight(abstract_txt:items in 1136) [ClassicSimilarity], result of:
           0.20027743 = score(doc=1136,freq=3.0), product of:
             0.26530248 = queryWeight, product of:
               2.8590608 = boost
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.01663328 = queryNorm
             0.75490224 = fieldWeight in 1136, product of:
               1.7320508 = tf(freq=3.0), with freq of:
                 3.0 = termFreq=3.0
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
         0.15794784 = weight(abstract_txt:recommendations in 1136) [ClassicSimilarity], result of:
           0.15794784 = score(doc=1136,freq=1.0), product of:
             0.3518354 = queryWeight, product of:
               3.681099 = boost
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.01663328 = queryNorm
             0.44892538 = fieldWeight in 1136, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.078125 = fieldNorm(doc=1136)
       0.24 = coord(6/25)
   

2. Lai, C,-H.: Applying knowledge flow mining to group recommendation methods for task-based groups (2015) 0.17

   0.1673246 = sum of:
     0.1673246 = product of:
       0.5975879 = sum of:
         0.023641547 = weight(abstract_txt:problem in 1667) [ClassicSimilarity], result of:
           0.023641547 = score(doc=1667,freq=1.0), product of:
             0.08480231 = queryWeight, product of:
               1.1429877 = boost
               4.460548 = idf(docFreq=1388, maxDocs=44218)
               0.01663328 = queryNorm
             0.27878425 = fieldWeight in 1667, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.460548 = idf(docFreq=1388, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.019678995 = weight(abstract_txt:different in 1667) [ClassicSimilarity], result of:
           0.019678995 = score(doc=1667,freq=1.0), product of:
             0.08589926 = queryWeight, product of:
               1.4088932 = boost
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.01663328 = queryNorm
             0.22909386 = fieldWeight in 1667, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.054149397 = weight(abstract_txt:collaborative in 1667) [ClassicSimilarity], result of:
           0.054149397 = score(doc=1667,freq=1.0), product of:
             0.1473509 = queryWeight, product of:
               1.5066562 = boost
               5.879776 = idf(docFreq=335, maxDocs=44218)
               0.01663328 = queryNorm
             0.367486 = fieldWeight in 1667, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.879776 = idf(docFreq=335, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.10527554 = weight(abstract_txt:filtering in 1667) [ClassicSimilarity], result of:
           0.10527554 = score(doc=1667,freq=2.0), product of:
             0.18217917 = queryWeight, product of:
               1.6752788 = boost
               6.537832 = idf(docFreq=173, maxDocs=44218)
               0.01663328 = queryNorm
             0.57786816 = fieldWeight in 1667, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               6.537832 = idf(docFreq=173, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.081601076 = weight(abstract_txt:recommended in 1667) [ClassicSimilarity], result of:
           0.081601076 = score(doc=1667,freq=1.0), product of:
             0.19368128 = queryWeight, product of:
               1.7273549 = boost
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.01663328 = queryNorm
             0.42131627 = fieldWeight in 1667, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.18688305 = weight(abstract_txt:recommendation in 1667) [ClassicSimilarity], result of:
           0.18688305 = score(doc=1667,freq=4.0), product of:
             0.2119908 = queryWeight, product of:
               1.8071585 = boost
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.01663328 = queryNorm
             0.88156205 = fieldWeight in 1667, product of:
               2.0 = tf(freq=4.0), with freq of:
                 4.0 = termFreq=4.0
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
         0.12635827 = weight(abstract_txt:recommendations in 1667) [ClassicSimilarity], result of:
           0.12635827 = score(doc=1667,freq=1.0), product of:
             0.3518354 = queryWeight, product of:
               3.681099 = boost
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.01663328 = queryNorm
             0.3591403 = fieldWeight in 1667, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.0625 = fieldNorm(doc=1667)
       0.28 = coord(7/25)
   

3. Soledad Pera, M.; Ng, Y.-K.: Recommending books to be exchanged online in the absence of wish lists (2018) 0.13

   0.12918577 = sum of:
     0.12918577 = product of:
       0.6459288 = sum of:
         0.054789394 = weight(abstract_txt:them in 4182) [ClassicSimilarity], result of:
           0.054789394 = score(doc=4182,freq=3.0), product of:
             0.11787221 = queryWeight, product of:
               1.6503994 = boost
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.01663328 = queryNorm
             0.4648203 = fieldWeight in 4182, product of:
               1.7320508 = tf(freq=3.0), with freq of:
                 3.0 = termFreq=3.0
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.0625 = fieldNorm(doc=4182)
         0.11540134 = weight(abstract_txt:recommended in 4182) [ClassicSimilarity], result of:
           0.11540134 = score(doc=4182,freq=2.0), product of:
             0.19368128 = queryWeight, product of:
               1.7273549 = boost
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.01663328 = queryNorm
             0.59583116 = fieldWeight in 4182, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               6.7410603 = idf(docFreq=141, maxDocs=44218)
               0.0625 = fieldNorm(doc=4182)
         0.16184546 = weight(abstract_txt:recommendation in 4182) [ClassicSimilarity], result of:
           0.16184546 = score(doc=4182,freq=3.0), product of:
             0.2119908 = queryWeight, product of:
               1.8071585 = boost
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.01663328 = queryNorm
             0.7634551 = fieldWeight in 4182, product of:
               1.7320508 = tf(freq=3.0), with freq of:
                 3.0 = termFreq=3.0
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.0625 = fieldNorm(doc=4182)
         0.10704695 = weight(abstract_txt:ratings in 4182) [ClassicSimilarity], result of:
           0.10704695 = score(doc=4182,freq=1.0), product of:
             0.232099 = queryWeight, product of:
               1.8909253 = boost
               7.3793993 = idf(docFreq=74, maxDocs=44218)
               0.01663328 = queryNorm
             0.46121246 = fieldWeight in 4182, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               7.3793993 = idf(docFreq=74, maxDocs=44218)
               0.0625 = fieldNorm(doc=4182)
         0.20684563 = weight(abstract_txt:items in 4182) [ClassicSimilarity], result of:
           0.20684563 = score(doc=4182,freq=5.0), product of:
             0.26530248 = queryWeight, product of:
               2.8590608 = boost
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.01663328 = queryNorm
             0.7796596 = fieldWeight in 4182, product of:
               2.236068 = tf(freq=5.0), with freq of:
                 5.0 = termFreq=5.0
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.0625 = fieldNorm(doc=4182)
       0.2 = coord(5/25)
   

4. Polat, H.; Du, W.: Privacy-preserving top-N recommendation on distributed data (2008) 0.11

   0.11199909 = sum of:
     0.11199909 = product of:
       0.5599954 = sum of:
         0.024598742 = weight(abstract_txt:different in 1864) [ClassicSimilarity], result of:
           0.024598742 = score(doc=1864,freq=1.0), product of:
             0.08589926 = queryWeight, product of:
               1.4088932 = boost
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.01663328 = queryNorm
             0.28636733 = fieldWeight in 1864, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               3.6655018 = idf(docFreq=3075, maxDocs=44218)
               0.078125 = fieldNorm(doc=1864)
         0.067686744 = weight(abstract_txt:collaborative in 1864) [ClassicSimilarity], result of:
           0.067686744 = score(doc=1864,freq=1.0), product of:
             0.1473509 = queryWeight, product of:
               1.5066562 = boost
               5.879776 = idf(docFreq=335, maxDocs=44218)
               0.01663328 = queryNorm
             0.4593575 = fieldWeight in 1864, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.879776 = idf(docFreq=335, maxDocs=44218)
               0.078125 = fieldNorm(doc=1864)
         0.13159443 = weight(abstract_txt:filtering in 1864) [ClassicSimilarity], result of:
           0.13159443 = score(doc=1864,freq=2.0), product of:
             0.18217917 = queryWeight, product of:
               1.6752788 = boost
               6.537832 = idf(docFreq=173, maxDocs=44218)
               0.01663328 = queryNorm
             0.7223352 = fieldWeight in 1864, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               6.537832 = idf(docFreq=173, maxDocs=44218)
               0.078125 = fieldNorm(doc=1864)
         0.20230682 = weight(abstract_txt:recommendation in 1864) [ClassicSimilarity], result of:
           0.20230682 = score(doc=1864,freq=3.0), product of:
             0.2119908 = queryWeight, product of:
               1.8071585 = boost
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.01663328 = queryNorm
             0.9543189 = fieldWeight in 1864, product of:
               1.7320508 = tf(freq=3.0), with freq of:
                 3.0 = termFreq=3.0
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.078125 = fieldNorm(doc=1864)
         0.13380869 = weight(abstract_txt:ratings in 1864) [ClassicSimilarity], result of:
           0.13380869 = score(doc=1864,freq=1.0), product of:
             0.232099 = queryWeight, product of:
               1.8909253 = boost
               7.3793993 = idf(docFreq=74, maxDocs=44218)
               0.01663328 = queryNorm
             0.57651556 = fieldWeight in 1864, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               7.3793993 = idf(docFreq=74, maxDocs=44218)
               0.078125 = fieldNorm(doc=1864)
       0.2 = coord(5/25)
   

5. Xu, X.; Dutta, K.; Datta, A.; Ge, C.: Identifying functional aspects from user reviews for functionality-based mobile app recommendation (2018) 0.11

   0.111072004 = sum of:
     0.111072004 = product of:
       0.55536 = sum of:
         0.023641547 = weight(abstract_txt:problem in 4038) [ClassicSimilarity], result of:
           0.023641547 = score(doc=4038,freq=1.0), product of:
             0.08480231 = queryWeight, product of:
               1.1429877 = boost
               4.460548 = idf(docFreq=1388, maxDocs=44218)
               0.01663328 = queryNorm
             0.27878425 = fieldWeight in 4038, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.460548 = idf(docFreq=1388, maxDocs=44218)
               0.0625 = fieldNorm(doc=4038)
         0.031632673 = weight(abstract_txt:them in 4038) [ClassicSimilarity], result of:
           0.031632673 = score(doc=4038,freq=1.0), product of:
             0.11787221 = queryWeight, product of:
               1.6503994 = boost
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.01663328 = queryNorm
             0.26836413 = fieldWeight in 4038, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.293826 = idf(docFreq=1640, maxDocs=44218)
               0.0625 = fieldNorm(doc=4038)
         0.22888407 = weight(abstract_txt:recommendation in 4038) [ClassicSimilarity], result of:
           0.22888407 = score(doc=4038,freq=6.0), product of:
             0.2119908 = queryWeight, product of:
               1.8071585 = boost
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.01663328 = queryNorm
             1.0796887 = fieldWeight in 4038, product of:
               2.4494898 = tf(freq=6.0), with freq of:
                 6.0 = termFreq=6.0
               7.0524964 = idf(docFreq=103, maxDocs=44218)
               0.0625 = fieldNorm(doc=4038)
         0.09250418 = weight(abstract_txt:items in 4038) [ClassicSimilarity], result of:
           0.09250418 = score(doc=4038,freq=1.0), product of:
             0.26530248 = queryWeight, product of:
               2.8590608 = boost
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.01663328 = queryNorm
             0.3486744 = fieldWeight in 4038, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.57879 = idf(docFreq=453, maxDocs=44218)
               0.0625 = fieldNorm(doc=4038)
         0.17869757 = weight(abstract_txt:recommendations in 4038) [ClassicSimilarity], result of:
           0.17869757 = score(doc=4038,freq=2.0), product of:
             0.3518354 = queryWeight, product of:
               3.681099 = boost
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.01663328 = queryNorm
             0.5079011 = fieldWeight in 4038, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               5.746245 = idf(docFreq=383, maxDocs=44218)
               0.0625 = fieldNorm(doc=4038)
       0.2 = coord(5/25)