Graph neural networks in node classification: survey and evaluation View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2021-11-02

AUTHORS

Shunxin Xiao, Shiping Wang, Yuanfei Dai, Wenzhong Guo

ABSTRACT

Neural networks have been proved efficient in improving many machine learning tasks such as convolutional neural networks and recurrent neural networks for computer vision and natural language processing, respectively. However, the inputs of these deep learning paradigms all belong to the type of Euclidean structure, e.g., images or texts. It is difficult to directly apply these neural networks to graph-based applications such as node classification since graph is a typical non-Euclidean structure in machine learning domain. Graph neural networks are designed to deal with the particular graph-based input and have received great developments because of more and more research attention. In this paper, we provide a comprehensive review about applying graph neural networks to the node classification task. First, the state-of-the-art methods are discussed and divided into three main categories: convolutional mechanism, attention mechanism and autoencoder mechanism. Afterward, extensive comparative experiments are conducted on several benchmark datasets, including citation networks and co-author networks, to compare the performance of different methods with diverse evaluation metrics. Finally, several suggestions are provided for future research based on the experimental results. More... »

PAGES

4

References to SciGraph publications

  • 2016-09-04. Multilabel Classification on Heterogeneous Graphs with Gaussian Embeddings in MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES
  • 2013. Mitosis Detection in Breast Cancer Histology Images with Deep Neural Networks in MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION – MICCAI 2013
  • <error retrieving object. in <ERROR RETRIEVING OBJECT
  • 2009-02-04. Complex brain networks: graph theoretical analysis of structural and functional systems in NATURE REVIEWS NEUROSCIENCE
  • 2016-01-27. Mastering the game of Go with deep neural networks and tree search in NATURE
  • Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/s00138-021-01251-0

    DOI

    http://dx.doi.org/10.1007/s00138-021-01251-0

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1142349239


    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/2008/0801", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Artificial Intelligence and Image Processing", 
            "type": "DefinedTerm"
          }
        ], 
        "author": [
          {
            "affiliation": {
              "alternateName": "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China", 
              "id": "http://www.grid.ac/institutes/grid.411604.6", 
              "name": [
                "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Xiao", 
            "givenName": "Shunxin", 
            "id": "sg:person.012177654000.11", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012177654000.11"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China", 
              "id": "http://www.grid.ac/institutes/grid.411604.6", 
              "name": [
                "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Wang", 
            "givenName": "Shiping", 
            "id": "sg:person.07661256461.05", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07661256461.05"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China", 
              "id": "http://www.grid.ac/institutes/grid.411604.6", 
              "name": [
                "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Dai", 
            "givenName": "Yuanfei", 
            "id": "sg:person.016026651040.54", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016026651040.54"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China", 
              "id": "http://www.grid.ac/institutes/grid.411604.6", 
              "name": [
                "College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Guo", 
            "givenName": "Wenzhong", 
            "id": "sg:person.015321651431.28", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015321651431.28"
            ], 
            "type": "Person"
          }
        ], 
        "citation": [
          {
            "id": "sg:pub.10.1007/978-3-319-46227-1_38", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1039495990", 
              "https://doi.org/10.1007/978-3-319-46227-1_38"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1038/nrn2575", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1004953014", 
              "https://doi.org/10.1038/nrn2575"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-3-7908-2604-3_16", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1017229575", 
              "https://doi.org/10.1007/978-3-7908-2604-3_16"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-3-642-40763-5_51", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1014968475", 
              "https://doi.org/10.1007/978-3-642-40763-5_51"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1038/nature16961", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1039427823", 
              "https://doi.org/10.1038/nature16961"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "2021-11-02", 
        "datePublishedReg": "2021-11-02", 
        "description": "Neural networks have been proved efficient in improving many machine learning tasks such as convolutional neural networks and recurrent neural networks for computer vision and natural language processing, respectively. However, the inputs of these deep learning paradigms all belong to the type of Euclidean structure, e.g., images or texts. It is difficult to directly apply these neural networks to graph-based applications such as node classification since graph is a typical non-Euclidean structure in machine learning domain. Graph neural networks are designed to deal with the particular graph-based input and have received great developments because of more and more research attention. In this paper, we provide a comprehensive review about applying graph neural networks to the node classification task. First, the state-of-the-art methods are discussed and divided into three main categories: convolutional mechanism, attention mechanism and autoencoder mechanism. Afterward, extensive comparative experiments are conducted on several benchmark datasets, including citation networks and co-author networks, to compare the performance of different methods with diverse evaluation metrics. Finally, several suggestions are provided for future research based on the experimental results.", 
        "genre": "article", 
        "id": "sg:pub.10.1007/s00138-021-01251-0", 
        "inLanguage": "en", 
        "isAccessibleForFree": true, 
        "isFundedItemOf": [
          {
            "id": "sg:grant.8306250", 
            "type": "MonetaryGrant"
          }, 
          {
            "id": "sg:grant.8299380", 
            "type": "MonetaryGrant"
          }
        ], 
        "isPartOf": [
          {
            "id": "sg:journal.1045266", 
            "issn": [
              "0932-8092", 
              "1432-1769"
            ], 
            "name": "Machine Vision and Applications", 
            "publisher": "Springer Nature", 
            "type": "Periodical"
          }, 
          {
            "issueNumber": "1", 
            "type": "PublicationIssue"
          }, 
          {
            "type": "PublicationVolume", 
            "volumeNumber": "33"
          }
        ], 
        "keywords": [
          "graph neural network", 
          "neural network", 
          "node classification", 
          "deep learning paradigm", 
          "graph-based applications", 
          "convolutional neural network", 
          "natural language processing", 
          "node classification task", 
          "recurrent neural network", 
          "non-Euclidean structure", 
          "Extensive comparative experiments", 
          "computer vision", 
          "classification task", 
          "attention mechanism", 
          "benchmark datasets", 
          "co-author network", 
          "language processing", 
          "art methods", 
          "evaluation metrics", 
          "learning paradigm", 
          "network", 
          "comparative experiments", 
          "citation network", 
          "Euclidean structure", 
          "more research attention", 
          "experimental results", 
          "machine", 
          "task", 
          "classification", 
          "research attention", 
          "great development", 
          "dataset", 
          "graph", 
          "input", 
          "metrics", 
          "images", 
          "vision", 
          "paradigm", 
          "processing", 
          "main categories", 
          "text", 
          "applications", 
          "different methods", 
          "performance", 
          "method", 
          "domain", 
          "comprehensive review", 
          "experiments", 
          "future research", 
          "research", 
          "evaluation", 
          "attention", 
          "categories", 
          "structure", 
          "development", 
          "state", 
          "results", 
          "mechanism", 
          "suggestions", 
          "types", 
          "survey", 
          "review", 
          "paper", 
          "typical non-Euclidean structure", 
          "particular graph-based input", 
          "graph-based input", 
          "convolutional mechanism", 
          "autoencoder mechanism", 
          "diverse evaluation metrics"
        ], 
        "name": "Graph neural networks in node classification: survey and evaluation", 
        "pagination": "4", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1142349239"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/s00138-021-01251-0"
            ]
          }
        ], 
        "sameAs": [
          "https://doi.org/10.1007/s00138-021-01251-0", 
          "https://app.dimensions.ai/details/publication/pub.1142349239"
        ], 
        "sdDataset": "articles", 
        "sdDatePublished": "2022-01-01T19:04", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/article/article_913.jsonl", 
        "type": "ScholarlyArticle", 
        "url": "https://doi.org/10.1007/s00138-021-01251-0"
      }
    ]
     

    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/s00138-021-01251-0'

    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/s00138-021-01251-0'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s00138-021-01251-0'

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

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s00138-021-01251-0'


     

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

    172 TRIPLES      22 PREDICATES      99 URIs      86 LITERALS      6 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/s00138-021-01251-0 schema:about anzsrc-for:08
    2 anzsrc-for:0801
    3 schema:author N68c8ca7f44fb4bd4970306361cfd36f3
    4 schema:citation sg:pub.10.1007/978-3-319-46227-1_38
    5 sg:pub.10.1007/978-3-642-40763-5_51
    6 sg:pub.10.1007/978-3-7908-2604-3_16
    7 sg:pub.10.1038/nature16961
    8 sg:pub.10.1038/nrn2575
    9 schema:datePublished 2021-11-02
    10 schema:datePublishedReg 2021-11-02
    11 schema:description Neural networks have been proved efficient in improving many machine learning tasks such as convolutional neural networks and recurrent neural networks for computer vision and natural language processing, respectively. However, the inputs of these deep learning paradigms all belong to the type of Euclidean structure, e.g., images or texts. It is difficult to directly apply these neural networks to graph-based applications such as node classification since graph is a typical non-Euclidean structure in machine learning domain. Graph neural networks are designed to deal with the particular graph-based input and have received great developments because of more and more research attention. In this paper, we provide a comprehensive review about applying graph neural networks to the node classification task. First, the state-of-the-art methods are discussed and divided into three main categories: convolutional mechanism, attention mechanism and autoencoder mechanism. Afterward, extensive comparative experiments are conducted on several benchmark datasets, including citation networks and co-author networks, to compare the performance of different methods with diverse evaluation metrics. Finally, several suggestions are provided for future research based on the experimental results.
    12 schema:genre article
    13 schema:inLanguage en
    14 schema:isAccessibleForFree true
    15 schema:isPartOf N34bda4b4edf94c4eba4644a85c48c62f
    16 Na7aebc3f3f4149b6a6ade735ac0d5b88
    17 sg:journal.1045266
    18 schema:keywords Euclidean structure
    19 Extensive comparative experiments
    20 applications
    21 art methods
    22 attention
    23 attention mechanism
    24 autoencoder mechanism
    25 benchmark datasets
    26 categories
    27 citation network
    28 classification
    29 classification task
    30 co-author network
    31 comparative experiments
    32 comprehensive review
    33 computer vision
    34 convolutional mechanism
    35 convolutional neural network
    36 dataset
    37 deep learning paradigm
    38 development
    39 different methods
    40 diverse evaluation metrics
    41 domain
    42 evaluation
    43 evaluation metrics
    44 experimental results
    45 experiments
    46 future research
    47 graph
    48 graph neural network
    49 graph-based applications
    50 graph-based input
    51 great development
    52 images
    53 input
    54 language processing
    55 learning paradigm
    56 machine
    57 main categories
    58 mechanism
    59 method
    60 metrics
    61 more research attention
    62 natural language processing
    63 network
    64 neural network
    65 node classification
    66 node classification task
    67 non-Euclidean structure
    68 paper
    69 paradigm
    70 particular graph-based input
    71 performance
    72 processing
    73 recurrent neural network
    74 research
    75 research attention
    76 results
    77 review
    78 state
    79 structure
    80 suggestions
    81 survey
    82 task
    83 text
    84 types
    85 typical non-Euclidean structure
    86 vision
    87 schema:name Graph neural networks in node classification: survey and evaluation
    88 schema:pagination 4
    89 schema:productId N43603238b40c4892b905e3b2c134fc92
    90 Nb922cc366be042f78952b6cdcc63e37e
    91 schema:sameAs https://app.dimensions.ai/details/publication/pub.1142349239
    92 https://doi.org/10.1007/s00138-021-01251-0
    93 schema:sdDatePublished 2022-01-01T19:04
    94 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    95 schema:sdPublisher Nada1d065f7734073861670ef960cd434
    96 schema:url https://doi.org/10.1007/s00138-021-01251-0
    97 sgo:license sg:explorer/license/
    98 sgo:sdDataset articles
    99 rdf:type schema:ScholarlyArticle
    100 N27df5561a5bf4c07b8b077a3d5a2dfa5 rdf:first sg:person.07661256461.05
    101 rdf:rest Nda3975ea8f4d489aab788d5773416675
    102 N34bda4b4edf94c4eba4644a85c48c62f schema:volumeNumber 33
    103 rdf:type schema:PublicationVolume
    104 N43603238b40c4892b905e3b2c134fc92 schema:name doi
    105 schema:value 10.1007/s00138-021-01251-0
    106 rdf:type schema:PropertyValue
    107 N68c8ca7f44fb4bd4970306361cfd36f3 rdf:first sg:person.012177654000.11
    108 rdf:rest N27df5561a5bf4c07b8b077a3d5a2dfa5
    109 Na7aebc3f3f4149b6a6ade735ac0d5b88 schema:issueNumber 1
    110 rdf:type schema:PublicationIssue
    111 Nada1d065f7734073861670ef960cd434 schema:name Springer Nature - SN SciGraph project
    112 rdf:type schema:Organization
    113 Nb922cc366be042f78952b6cdcc63e37e schema:name dimensions_id
    114 schema:value pub.1142349239
    115 rdf:type schema:PropertyValue
    116 Nda3975ea8f4d489aab788d5773416675 rdf:first sg:person.016026651040.54
    117 rdf:rest Nf0ce4bb4d5e04b2583d6f3c3015ff3f9
    118 Nf0ce4bb4d5e04b2583d6f3c3015ff3f9 rdf:first sg:person.015321651431.28
    119 rdf:rest rdf:nil
    120 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
    121 schema:name Information and Computing Sciences
    122 rdf:type schema:DefinedTerm
    123 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
    124 schema:name Artificial Intelligence and Image Processing
    125 rdf:type schema:DefinedTerm
    126 sg:grant.8299380 http://pending.schema.org/fundedItem sg:pub.10.1007/s00138-021-01251-0
    127 rdf:type schema:MonetaryGrant
    128 sg:grant.8306250 http://pending.schema.org/fundedItem sg:pub.10.1007/s00138-021-01251-0
    129 rdf:type schema:MonetaryGrant
    130 sg:journal.1045266 schema:issn 0932-8092
    131 1432-1769
    132 schema:name Machine Vision and Applications
    133 schema:publisher Springer Nature
    134 rdf:type schema:Periodical
    135 sg:person.012177654000.11 schema:affiliation grid-institutes:grid.411604.6
    136 schema:familyName Xiao
    137 schema:givenName Shunxin
    138 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012177654000.11
    139 rdf:type schema:Person
    140 sg:person.015321651431.28 schema:affiliation grid-institutes:grid.411604.6
    141 schema:familyName Guo
    142 schema:givenName Wenzhong
    143 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015321651431.28
    144 rdf:type schema:Person
    145 sg:person.016026651040.54 schema:affiliation grid-institutes:grid.411604.6
    146 schema:familyName Dai
    147 schema:givenName Yuanfei
    148 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016026651040.54
    149 rdf:type schema:Person
    150 sg:person.07661256461.05 schema:affiliation grid-institutes:grid.411604.6
    151 schema:familyName Wang
    152 schema:givenName Shiping
    153 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07661256461.05
    154 rdf:type schema:Person
    155 sg:pub.10.1007/978-3-319-46227-1_38 schema:sameAs https://app.dimensions.ai/details/publication/pub.1039495990
    156 https://doi.org/10.1007/978-3-319-46227-1_38
    157 rdf:type schema:CreativeWork
    158 sg:pub.10.1007/978-3-642-40763-5_51 schema:sameAs https://app.dimensions.ai/details/publication/pub.1014968475
    159 https://doi.org/10.1007/978-3-642-40763-5_51
    160 rdf:type schema:CreativeWork
    161 sg:pub.10.1007/978-3-7908-2604-3_16 schema:sameAs https://app.dimensions.ai/details/publication/pub.1017229575
    162 https://doi.org/10.1007/978-3-7908-2604-3_16
    163 rdf:type schema:CreativeWork
    164 sg:pub.10.1038/nature16961 schema:sameAs https://app.dimensions.ai/details/publication/pub.1039427823
    165 https://doi.org/10.1038/nature16961
    166 rdf:type schema:CreativeWork
    167 sg:pub.10.1038/nrn2575 schema:sameAs https://app.dimensions.ai/details/publication/pub.1004953014
    168 https://doi.org/10.1038/nrn2575
    169 rdf:type schema:CreativeWork
    170 grid-institutes:grid.411604.6 schema:alternateName College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China
    171 schema:name College of Mathematics and Computer Sciences, Fuzhou University, Fuzhou, China
    172 rdf:type schema:Organization
     




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


    ...