Optimizing the recency-relevance-diversity trade-offs in non-personalized news recommendations View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2019-02-27

AUTHORS

Abhijnan Chakraborty, Saptarshi Ghosh, Niloy Ganguly, Krishna P. Gummadi

ABSTRACT

Online news media sites are emerging as the primary source of news for a large number of users. Due to a large number of stories being published in these media sites, users usually rely on news recommendation systems to find important news. In this work, we focus on automatically recommending news stories to all users of such media websites, where the selection is not influenced by a particular user’s news reading habit. When recommending news stories in such non-personalized manner, there are three basic metrics of interest—recency, importance (analogous to relevance in personalized recommendation) and diversity of the recommended news. Ideally, recommender systems should recommend the most important stories soon after they are published. However, the importance of a story only becomes evident as the story ages, thereby creating a tension between recency and importance. A systematic analysis of popular recommendation strategies in use today reveals that they lead to poor trade-offs between recency and importance in practice. So, in this paper, we propose a new recommendation strategy (called Highest Future-Impact) which attempts to optimize on both the axes. To implement our proposed strategy in practice, we propose two approaches to predict the future-impact of news stories, by using crowd-sourced popularity signals and by observing editorial selection in past news data. Finally, we propose approaches to inculcate diversity in recommended news which can maintain a balanced proportion of news from different news sections. Evaluations over real-world news datasets show that our implementations achieve good performance in recommending news stories. More... »

PAGES

447-475

References to SciGraph publications

  • 2007. Rhythms of Social Interaction: Messaging Within a Massive Online Network in COMMUNITIES AND TECHNOLOGIES 2007
  • 2007-08-08. A note on Platt’s probabilistic outputs for support vector machines in MACHINE LEARNING
  • 2011-09-23. Personalized News Recommendation: A Review and an Experimental Investigation in JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY
  • 2013-07-09. Multi-objective Optimization in SEARCH METHODOLOGIES
  • Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/s10791-019-09351-2

    DOI

    http://dx.doi.org/10.1007/s10791-019-09351-2

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1112433937


    Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
    Incoming Citations Browse incoming citations for this publication using opencitations.net

    JSON-LD is the canonical representation for SciGraph data.

    TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

    [
      {
        "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
        "about": [
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Information and Computing Sciences", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0804", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Data Format", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0806", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Information Systems", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0807", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Library and Information Studies", 
            "type": "DefinedTerm"
          }
        ], 
        "author": [
          {
            "affiliation": {
              "alternateName": "Indian Institute of Technology Kharagpur, Kharagpur, India", 
              "id": "http://www.grid.ac/institutes/grid.429017.9", 
              "name": [
                "Max Planck Institute for Software Systems, Saarbr\u00fccken, Germany", 
                "Indian Institute of Technology Kharagpur, Kharagpur, India"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Chakraborty", 
            "givenName": "Abhijnan", 
            "id": "sg:person.010072524745.28", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010072524745.28"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Indian Institute of Technology Kharagpur, Kharagpur, India", 
              "id": "http://www.grid.ac/institutes/grid.429017.9", 
              "name": [
                "Indian Institute of Technology Kharagpur, Kharagpur, India"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Ghosh", 
            "givenName": "Saptarshi", 
            "id": "sg:person.012073064277.60", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012073064277.60"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Indian Institute of Technology Kharagpur, Kharagpur, India", 
              "id": "http://www.grid.ac/institutes/grid.429017.9", 
              "name": [
                "Indian Institute of Technology Kharagpur, Kharagpur, India"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Ganguly", 
            "givenName": "Niloy", 
            "id": "sg:person.014562044325.07", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014562044325.07"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Max Planck Institute for Software Systems, Saarbr\u00fccken, Germany", 
              "id": "http://www.grid.ac/institutes/grid.469860.5", 
              "name": [
                "Max Planck Institute for Software Systems, Saarbr\u00fccken, Germany"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Gummadi", 
            "givenName": "Krishna P.", 
            "id": "sg:person.012574113025.70", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012574113025.70"
            ], 
            "type": "Person"
          }
        ], 
        "citation": [
          {
            "id": "sg:pub.10.1007/978-1-4614-6940-7_15", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1026110769", 
              "https://doi.org/10.1007/978-1-4614-6940-7_15"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-1-84628-905-7_3", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1022061902", 
              "https://doi.org/10.1007/978-1-84628-905-7_3"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/s11390-011-0175-2", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1043638766", 
              "https://doi.org/10.1007/s11390-011-0175-2"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/s10994-007-5018-6", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1039556605", 
              "https://doi.org/10.1007/s10994-007-5018-6"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "2019-02-27", 
        "datePublishedReg": "2019-02-27", 
        "description": "Online news media sites are emerging as the primary source of news for a large number of users. Due to a large number of stories being published in these media sites, users usually rely on news recommendation systems to find important news. In this work, we focus on automatically recommending news stories to all users of such media websites, where the selection is not influenced by a particular user\u2019s news reading habit. When recommending news stories in such non-personalized manner, there are three basic metrics of interest\u2014recency, importance (analogous to relevance in personalized recommendation) and diversity of the recommended news. Ideally, recommender systems should recommend the most important stories soon after they are published. However, the importance of a story only becomes evident as the story ages, thereby creating a tension between recency and importance. A systematic analysis of popular recommendation strategies in use today reveals that they lead to poor trade-offs between recency and importance in practice. So, in this paper, we propose a new recommendation strategy (called Highest Future-Impact) which attempts to optimize on both the axes. To implement our proposed strategy in practice, we propose two approaches to predict the future-impact of news stories, by using crowd-sourced popularity signals and by observing editorial selection in past news data. Finally, we propose approaches to inculcate diversity in recommended news which can maintain a balanced proportion of news from different news sections. Evaluations over real-world news datasets show that our implementations achieve good performance in recommending news stories.", 
        "genre": "article", 
        "id": "sg:pub.10.1007/s10791-019-09351-2", 
        "inLanguage": "en", 
        "isAccessibleForFree": true, 
        "isFundedItemOf": [
          {
            "id": "sg:grant.7613196", 
            "type": "MonetaryGrant"
          }
        ], 
        "isPartOf": [
          {
            "id": "sg:journal.1023664", 
            "issn": [
              "1386-4564", 
              "1573-7659"
            ], 
            "name": "Information Retrieval Journal", 
            "publisher": "Springer Nature", 
            "type": "Periodical"
          }, 
          {
            "issueNumber": "5", 
            "type": "PublicationIssue"
          }, 
          {
            "type": "PublicationVolume", 
            "volumeNumber": "22"
          }
        ], 
        "keywords": [
          "new reading habits", 
          "news stories", 
          "media sites", 
          "news media sites", 
          "new recommendation strategy", 
          "editorial selection", 
          "reading habits", 
          "media websites", 
          "important stories", 
          "real-world news datasets", 
          "important news", 
          "story", 
          "news recommendation system", 
          "news", 
          "recommendation strategy", 
          "primary source", 
          "new section", 
          "practice", 
          "recommendation system", 
          "new dataset", 
          "users", 
          "diversity", 
          "websites", 
          "basic metrics", 
          "today", 
          "large number", 
          "better performance", 
          "tension", 
          "importance", 
          "use today", 
          "systematic analysis", 
          "work", 
          "habits", 
          "new recommendations", 
          "dataset", 
          "strategies", 
          "recency", 
          "system", 
          "metrics", 
          "implementation", 
          "approach", 
          "selection", 
          "new data", 
          "performance", 
          "source", 
          "paper", 
          "recommendations", 
          "number", 
          "sections", 
          "analysis", 
          "manner", 
          "data", 
          "age", 
          "sites", 
          "evaluation", 
          "balanced proportion", 
          "signals", 
          "axes", 
          "proportion", 
          "Online news media sites", 
          "such media websites", 
          "particular user\u2019s news reading habit", 
          "user\u2019s news reading habit", 
          "such non-personalized manner", 
          "non-personalized manner", 
          "story ages", 
          "popular recommendation strategies", 
          "crowd-sourced popularity signals", 
          "popularity signals", 
          "past news data", 
          "different news sections", 
          "non-personalized news recommendations"
        ], 
        "name": "Optimizing the recency-relevance-diversity trade-offs in non-personalized news recommendations", 
        "pagination": "447-475", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1112433937"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/s10791-019-09351-2"
            ]
          }
        ], 
        "sameAs": [
          "https://doi.org/10.1007/s10791-019-09351-2", 
          "https://app.dimensions.ai/details/publication/pub.1112433937"
        ], 
        "sdDataset": "articles", 
        "sdDatePublished": "2021-12-01T19:45", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com-springernature-scigraph/baseset/20211201/entities/gbq_results/article/article_805.jsonl", 
        "type": "ScholarlyArticle", 
        "url": "https://doi.org/10.1007/s10791-019-09351-2"
      }
    ]
     

    Download the RDF metadata as:  json-ld nt turtle xml License info

    HOW TO GET THIS DATA PROGRAMMATICALLY:

    JSON-LD is a popular format for linked data which is fully compatible with JSON.

    curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/s10791-019-09351-2'

    N-Triples is a line-based linked data format ideal for batch operations.

    curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/s10791-019-09351-2'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s10791-019-09351-2'

    RDF/XML is a standard XML format for linked data.

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s10791-019-09351-2'


     

    This table displays all metadata directly associated to this object as RDF triples.

    181 TRIPLES      22 PREDICATES      103 URIs      89 LITERALS      6 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/s10791-019-09351-2 schema:about anzsrc-for:08
    2 anzsrc-for:0804
    3 anzsrc-for:0806
    4 anzsrc-for:0807
    5 schema:author Na65cf656b781480184eee28f481eb03b
    6 schema:citation sg:pub.10.1007/978-1-4614-6940-7_15
    7 sg:pub.10.1007/978-1-84628-905-7_3
    8 sg:pub.10.1007/s10994-007-5018-6
    9 sg:pub.10.1007/s11390-011-0175-2
    10 schema:datePublished 2019-02-27
    11 schema:datePublishedReg 2019-02-27
    12 schema:description Online news media sites are emerging as the primary source of news for a large number of users. Due to a large number of stories being published in these media sites, users usually rely on news recommendation systems to find important news. In this work, we focus on automatically recommending news stories to all users of such media websites, where the selection is not influenced by a particular user’s news reading habit. When recommending news stories in such non-personalized manner, there are three basic metrics of interest—recency, importance (analogous to relevance in personalized recommendation) and diversity of the recommended news. Ideally, recommender systems should recommend the most important stories soon after they are published. However, the importance of a story only becomes evident as the story ages, thereby creating a tension between recency and importance. A systematic analysis of popular recommendation strategies in use today reveals that they lead to poor trade-offs between recency and importance in practice. So, in this paper, we propose a new recommendation strategy (called Highest Future-Impact) which attempts to optimize on both the axes. To implement our proposed strategy in practice, we propose two approaches to predict the future-impact of news stories, by using crowd-sourced popularity signals and by observing editorial selection in past news data. Finally, we propose approaches to inculcate diversity in recommended news which can maintain a balanced proportion of news from different news sections. Evaluations over real-world news datasets show that our implementations achieve good performance in recommending news stories.
    13 schema:genre article
    14 schema:inLanguage en
    15 schema:isAccessibleForFree true
    16 schema:isPartOf N2e935ceac31245cf8784ca41fdc3e8fd
    17 N5cfef5d435af412ab20d26c43e1961c5
    18 sg:journal.1023664
    19 schema:keywords Online news media sites
    20 age
    21 analysis
    22 approach
    23 axes
    24 balanced proportion
    25 basic metrics
    26 better performance
    27 crowd-sourced popularity signals
    28 data
    29 dataset
    30 different news sections
    31 diversity
    32 editorial selection
    33 evaluation
    34 habits
    35 implementation
    36 importance
    37 important news
    38 important stories
    39 large number
    40 manner
    41 media sites
    42 media websites
    43 metrics
    44 new data
    45 new dataset
    46 new reading habits
    47 new recommendation strategy
    48 new recommendations
    49 new section
    50 news
    51 news media sites
    52 news recommendation system
    53 news stories
    54 non-personalized manner
    55 non-personalized news recommendations
    56 number
    57 paper
    58 particular user’s news reading habit
    59 past news data
    60 performance
    61 popular recommendation strategies
    62 popularity signals
    63 practice
    64 primary source
    65 proportion
    66 reading habits
    67 real-world news datasets
    68 recency
    69 recommendation strategy
    70 recommendation system
    71 recommendations
    72 sections
    73 selection
    74 signals
    75 sites
    76 source
    77 story
    78 story ages
    79 strategies
    80 such media websites
    81 such non-personalized manner
    82 system
    83 systematic analysis
    84 tension
    85 today
    86 use today
    87 users
    88 user’s news reading habit
    89 websites
    90 work
    91 schema:name Optimizing the recency-relevance-diversity trade-offs in non-personalized news recommendations
    92 schema:pagination 447-475
    93 schema:productId N2061ff68a5a94f3f9f68bd2a5ac2aa2f
    94 N9eb9baffa46e4593abed1c73f87c56fd
    95 schema:sameAs https://app.dimensions.ai/details/publication/pub.1112433937
    96 https://doi.org/10.1007/s10791-019-09351-2
    97 schema:sdDatePublished 2021-12-01T19:45
    98 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    99 schema:sdPublisher N890856a713844a24be67cc06d0203793
    100 schema:url https://doi.org/10.1007/s10791-019-09351-2
    101 sgo:license sg:explorer/license/
    102 sgo:sdDataset articles
    103 rdf:type schema:ScholarlyArticle
    104 N2061ff68a5a94f3f9f68bd2a5ac2aa2f schema:name dimensions_id
    105 schema:value pub.1112433937
    106 rdf:type schema:PropertyValue
    107 N2e935ceac31245cf8784ca41fdc3e8fd schema:volumeNumber 22
    108 rdf:type schema:PublicationVolume
    109 N5c7a569d991f46eaaebbc7b7ee2a4f89 rdf:first sg:person.014562044325.07
    110 rdf:rest N9dd45550d1ba445ead8fd9bc683aaf6f
    111 N5cfef5d435af412ab20d26c43e1961c5 schema:issueNumber 5
    112 rdf:type schema:PublicationIssue
    113 N890856a713844a24be67cc06d0203793 schema:name Springer Nature - SN SciGraph project
    114 rdf:type schema:Organization
    115 N8c146864fe6049de9f0b035573199c5a rdf:first sg:person.012073064277.60
    116 rdf:rest N5c7a569d991f46eaaebbc7b7ee2a4f89
    117 N9dd45550d1ba445ead8fd9bc683aaf6f rdf:first sg:person.012574113025.70
    118 rdf:rest rdf:nil
    119 N9eb9baffa46e4593abed1c73f87c56fd schema:name doi
    120 schema:value 10.1007/s10791-019-09351-2
    121 rdf:type schema:PropertyValue
    122 Na65cf656b781480184eee28f481eb03b rdf:first sg:person.010072524745.28
    123 rdf:rest N8c146864fe6049de9f0b035573199c5a
    124 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
    125 schema:name Information and Computing Sciences
    126 rdf:type schema:DefinedTerm
    127 anzsrc-for:0804 schema:inDefinedTermSet anzsrc-for:
    128 schema:name Data Format
    129 rdf:type schema:DefinedTerm
    130 anzsrc-for:0806 schema:inDefinedTermSet anzsrc-for:
    131 schema:name Information Systems
    132 rdf:type schema:DefinedTerm
    133 anzsrc-for:0807 schema:inDefinedTermSet anzsrc-for:
    134 schema:name Library and Information Studies
    135 rdf:type schema:DefinedTerm
    136 sg:grant.7613196 http://pending.schema.org/fundedItem sg:pub.10.1007/s10791-019-09351-2
    137 rdf:type schema:MonetaryGrant
    138 sg:journal.1023664 schema:issn 1386-4564
    139 1573-7659
    140 schema:name Information Retrieval Journal
    141 schema:publisher Springer Nature
    142 rdf:type schema:Periodical
    143 sg:person.010072524745.28 schema:affiliation grid-institutes:grid.429017.9
    144 schema:familyName Chakraborty
    145 schema:givenName Abhijnan
    146 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010072524745.28
    147 rdf:type schema:Person
    148 sg:person.012073064277.60 schema:affiliation grid-institutes:grid.429017.9
    149 schema:familyName Ghosh
    150 schema:givenName Saptarshi
    151 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012073064277.60
    152 rdf:type schema:Person
    153 sg:person.012574113025.70 schema:affiliation grid-institutes:grid.469860.5
    154 schema:familyName Gummadi
    155 schema:givenName Krishna P.
    156 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012574113025.70
    157 rdf:type schema:Person
    158 sg:person.014562044325.07 schema:affiliation grid-institutes:grid.429017.9
    159 schema:familyName Ganguly
    160 schema:givenName Niloy
    161 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014562044325.07
    162 rdf:type schema:Person
    163 sg:pub.10.1007/978-1-4614-6940-7_15 schema:sameAs https://app.dimensions.ai/details/publication/pub.1026110769
    164 https://doi.org/10.1007/978-1-4614-6940-7_15
    165 rdf:type schema:CreativeWork
    166 sg:pub.10.1007/978-1-84628-905-7_3 schema:sameAs https://app.dimensions.ai/details/publication/pub.1022061902
    167 https://doi.org/10.1007/978-1-84628-905-7_3
    168 rdf:type schema:CreativeWork
    169 sg:pub.10.1007/s10994-007-5018-6 schema:sameAs https://app.dimensions.ai/details/publication/pub.1039556605
    170 https://doi.org/10.1007/s10994-007-5018-6
    171 rdf:type schema:CreativeWork
    172 sg:pub.10.1007/s11390-011-0175-2 schema:sameAs https://app.dimensions.ai/details/publication/pub.1043638766
    173 https://doi.org/10.1007/s11390-011-0175-2
    174 rdf:type schema:CreativeWork
    175 grid-institutes:grid.429017.9 schema:alternateName Indian Institute of Technology Kharagpur, Kharagpur, India
    176 schema:name Indian Institute of Technology Kharagpur, Kharagpur, India
    177 Max Planck Institute for Software Systems, Saarbrücken, Germany
    178 rdf:type schema:Organization
    179 grid-institutes:grid.469860.5 schema:alternateName Max Planck Institute for Software Systems, Saarbrücken, Germany
    180 schema:name Max Planck Institute for Software Systems, Saarbrücken, Germany
    181 rdf:type schema:Organization
     




    Preview window. Press ESC to close (or click here)


    ...