DETERRENT: Knowledge Guided Graph Attention Network for Detecting Healthcare Misinformation

KDD '20: The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining Virtual Event CA USA July, 2020, pp. 492-502, 2020.

Cited by: 0|Views85
EI
Weibo:
DETERRENT leverages additional information from a medical knowledge graph, to guide the article embedding with a graph attention network

Abstract:

To provide accurate and explainable misinformation detection, it is often useful to take an auxiliary source (e.g., social context and knowledge base) into consideration. Existing methods use social contexts such as users' engagements as complementary information to improve detection performance and derive explanations. However, due to th...More

Code:

Data:

0
Introduction
  • Fact

    Lower body mass index (BMI) is consistently associated with reduced type II diabetes risk, among people with varied family history, genetic risk factors and weight, according to a new study.

    Triples from KG (BMI, Diagnoses, Diabetes) (Family History, Causes, Diabetes)

    Misinformation Besides chemicals, cancer loves sugar.
  • Lower body mass index (BMI) is consistently associated with reduced type II diabetes risk, among people with varied family history, genetic risk factors and weight, according to a new study.
  • Misinformation Besides chemicals, cancer loves sugar.
  • The popularity of online social networks has promoted the growth of various applications and information, which enables users to browse and publish such information more freely.
  • Many studies [12, 32] have confirmed the existence and the spread of healthcare misinformation.
  • A study of three health social networking websites found that 54% of posts contained medical claims that are inaccurate or incomplete [38]
Highlights
  • Fact

    Lower body mass index (BMI) is consistently associated with reduced type II diabetes risk, among people with varied family history, genetic risk factors and weight, according to a new study.

    Triples from Knowledge graph (KG) (BMI, Diagnoses, Diabetes) (Family History, Causes, Diabetes)

    Misinformation Besides chemicals, cancer loves sugar
  • To address the above two issues, we propose a knowledge guided graph attention network that can better capture the crucial entities in news articles and guide the article embedding
  • We study a novel problem of explainable healthcare misinformation detection by leveraging medical knowledge graph to better capture the high-order relations between entities;
  • We propose a novel method DETERRENT, which characterizes multiple positive and negative relations in the medical knowledge graph under a relational graph attention network; and
  • We proposed DETERRENT, a knowledge guided graph attention network for misinformation detection in healthcare
  • DETERRENT leverages additional information from a medical knowledge graph, to guide the article embedding with a graph attention network
Methods
  • The authors' proposed framework consists of three components, which is shown in Figure 2: 1) an information propagation net, which propagates the knowledge between articles and nodes by preserving the structure of KG; 2) knowledge aware attention, which learns the weights of a node’s neighbors in KG and aggregates the information from the neighbors and an article’s contextual information to update its representation; 3) a prediction layer, which takes an article’s representation as input and outputs a predicted label.
  • The medical knowledge graph can provide medical evidence in healthcare misinformation detection.
  • To fully utilize the medical knowledge graph for healthcare misinformation detection, motivated by previous work [34, 37], the authors leverage inherent directional structure of the medical database to learn the entity embedding.
Conclusion
  • The authors proposed DETERRENT, a knowledge guided graph attention network for misinformation detection in healthcare.
  • DETERRENT leverages additional information from a medical knowledge graph, to guide the article embedding with a graph attention network.
  • DETERRENT has two limitations
  • It only leverages a knowledge graph, instead of other complementary information.
  • It does not consider the publishing time of an article.
  • Time intervals between posts can be considered to model misinformation diffusion
Summary
  • Introduction:

    Fact

    Lower body mass index (BMI) is consistently associated with reduced type II diabetes risk, among people with varied family history, genetic risk factors and weight, according to a new study.

    Triples from KG (BMI, Diagnoses, Diabetes) (Family History, Causes, Diabetes)

    Misinformation Besides chemicals, cancer loves sugar.
  • Lower body mass index (BMI) is consistently associated with reduced type II diabetes risk, among people with varied family history, genetic risk factors and weight, according to a new study.
  • Misinformation Besides chemicals, cancer loves sugar.
  • The popularity of online social networks has promoted the growth of various applications and information, which enables users to browse and publish such information more freely.
  • Many studies [12, 32] have confirmed the existence and the spread of healthcare misinformation.
  • A study of three health social networking websites found that 54% of posts contained medical claims that are inaccurate or incomplete [38]
  • Objectives:

    Where ˆ is the predicted value which indicates the probability of the article being fake.
  • The authors' goal is to minimize the cross-entropy loss:.
  • The authors present the experiments to evaluate the effectiveness of DETERRENT.
  • The authors aim to answer the following evaluation questions:
  • The authors present the experiments to evaluate the effectiveness of DETERRENT. the authors aim to answer the following evaluation questions:
  • Methods:

    The authors' proposed framework consists of three components, which is shown in Figure 2: 1) an information propagation net, which propagates the knowledge between articles and nodes by preserving the structure of KG; 2) knowledge aware attention, which learns the weights of a node’s neighbors in KG and aggregates the information from the neighbors and an article’s contextual information to update its representation; 3) a prediction layer, which takes an article’s representation as input and outputs a predicted label.
  • The medical knowledge graph can provide medical evidence in healthcare misinformation detection.
  • To fully utilize the medical knowledge graph for healthcare misinformation detection, motivated by previous work [34, 37], the authors leverage inherent directional structure of the medical database to learn the entity embedding.
  • Conclusion:

    The authors proposed DETERRENT, a knowledge guided graph attention network for misinformation detection in healthcare.
  • DETERRENT leverages additional information from a medical knowledge graph, to guide the article embedding with a graph attention network.
  • DETERRENT has two limitations
  • It only leverages a knowledge graph, instead of other complementary information.
  • It does not consider the publishing time of an article.
  • Time intervals between posts can be considered to model misinformation diffusion
Tables
  • Table1: Statistics of datasets
  • Table2: Performance Comparison on Diabetes and Cancer datasets. DETERRENT outperforms all state-of-the-art baselines including knowledge graph based and article contents based methods
  • Table3: Effects of the network depth
  • Table4: Ablation study of DETERRENT demonstrated the advantage of the attention mechanisms and modeling both positive and negative relations
  • Table5: The details of the parameters of DETERRENT
  • Table6: An example of the entity name consistency
  • Table7: Example triples extracted from specialized portals
  • Table8: Querying Examples of DETERRENT
Download tables as Excel
Related work
  • In this section, we briefly review two related topics: misinformation detection and graph neural networks.

    Misinformation Detection. Misinformation detection methods generally focus on using article contents and external sources. Article contents contain linguistic clues and visual factors that can differentiate the fake and real information. Linguistic features based methods check the consistency between the headlines and contents [4], or capture specific writing styles and sensational headlines that commonly occur in fake content [28]. Visual-based features can work with linguistic features to to identify fake images [42], and help to detect misinformation collectively [9, 13].
Funding
  • We thank Patrick Ernst and Gerhard Weikum for sharing KnowLife data with us, and Jason (Jiasheng) Zhang for his valuable feedback. This work was in part supported by NSF awards #1742702, #1820609, #1909702, #1915801, and #1934782
Reference
  • Hunt Allcott and Matthew Gentzkow. 2017. Social media and fake news in the 2016 election. Journal of economic perspectives 31, 2 (2017), 211–36.
    Google ScholarLocate open access versionFindings
  • Gabor Angeli, Melvin Jose Johnson Premkumar, and Christopher D Manning. 2015. Leveraging linguistic structure for open domain information extraction. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 344–354.
    Google ScholarLocate open access versionFindings
  • Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473 (2014).
    Findings
  • Gaurav Bhatt, Aman Sharma, Shivam Sharma, Ankush Nagpal, Balasubramanian Raman, and Ankush Mittal. 2018. Combining neural, statistical and external features for fake news stance identification. In Companion Proceedings of the The Web Conference 2018. 1353–1357.
    Google ScholarLocate open access versionFindings
  • Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating embeddings for modeling multi-relational data. In Advances in neural information processing systems. 2787–2795.
    Google ScholarFindings
  • Dan Busbridge, Dane Sherburn, Pietro Cavallo, and Nils Y Hammerla. 2019. Relational Graph Attention Networks. arXiv preprint arXiv:1904.05811 (2019).
    Findings
  • Giovanni Luca Ciampaglia, Prashant Shiralkar, Luis M Rocha, Johan Bollen, Filippo Menczer, and Alessandro Flammini. 2015. Computational fact checking from knowledge networks. PloS one 10, 6 (2015).
    Google ScholarLocate open access versionFindings
  • Limeng Cui, Kai Shu, Suhang Wang, Dongwon Lee, and Huan Liu. 2019. dEFEND: A System for Explainable Fake News Detection. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. 2961–2964.
    Google ScholarLocate open access versionFindings
  • Limeng Cui, Suhang Wang, and Dongwon Lee. 201SAME: sentiment-aware multi-modal embedding for detecting fake news. In Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. 41–48.
    Google ScholarLocate open access versionFindings
  • Tyler Derr, Yao Ma, and Jiliang Tang. 2018. Signed graph convolutional networks. In 2018 IEEE International Conference on Data Mining (ICDM). IEEE, 929–934.
    Google ScholarLocate open access versionFindings
  • Patrick Ernst, Amy Siu, and Gerhard Weikum. 2015. KnowLife: a versatile approach for constructing a large knowledge graph for biomedical sciences. BMC bioinformatics 16, 1 (2015), 157.
    Google ScholarLocate open access versionFindings
  • Gunther Eysenbach, John Powell, Oliver Kuss, and Eun-Ryoung Sa. 2002. Empirical studies assessing the quality of health information for consumers on the world wide web: a systematic review. Jama 287, 20 (2002), 2691–2700.
    Google ScholarLocate open access versionFindings
  • Han Guo, Juan Cao, Yazi Zhang, Junbo Guo, and Jintao Li. 2018. Rumor detection with hierarchical social attention network. In Proceedings of the 27th ACM International Conference on Information and Knowledge Management. 943–951.
    Google ScholarLocate open access versionFindings
  • Fritz Heider. 1946. Attitudes and cognitive organization. The Journal of psychology 21, 1 (1946), 107–112.
    Google ScholarLocate open access versionFindings
  • Viet-Phi Huynh and Paolo Papotti. 2019. A Benchmark for Fact Checking Algorithms Built on Knowledge Bases. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. 689–698.
    Google ScholarLocate open access versionFindings
  • Zhiwei Jin, Juan Cao, Yongdong Zhang, and Jiebo Luo. 20News verification by exploiting conflicting social viewpoints in microblogs. In Thirtieth AAAI conference on artificial intelligence.
    Google ScholarLocate open access versionFindings
  • Georgios Karagiannis, Immanuel Trummer, Saehan Jo, Shubham Khandelwal, Xuezhi Wang, and Cong Yu. 2019. Mining an" anti-knowledge base" from Wikipedia updates with applications to fact checking and beyond. Proceedings of the VLDB Endowment 13, 4 (2019), 561–573.
    Google ScholarLocate open access versionFindings
  • Yoon Kim. 2014. Convolutional Neural Networks for Sentence Classification. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). 1746–1751.
    Google ScholarLocate open access versionFindings
  • Diederik P Kingma and Jimmy Ba. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014).
    Findings
  • Thomas N Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 (2016).
    Findings
  • Quoc Le and Tomas Mikolov. 2014. Distributed representations of sentences and documents. In International conference on machine learning. 1188–1196.
    Google ScholarLocate open access versionFindings
  • Jure Leskovec, Daniel Huttenlocher, and Jon Kleinberg. 2010. Predicting positive and negative links in online social networks. In Proceedings of the 19th international conference on World wide web. 641–650.
    Google ScholarLocate open access versionFindings
  • Jure Leskovec, Daniel Huttenlocher, and Jon Kleinberg. 2010. Signed networks in social media. In Proceedings of the SIGCHI conference on human factors in computing systems. 1361–1370.
    Google ScholarLocate open access versionFindings
  • Hu Linmei, Tianchi Yang, Chuan Shi, Houye Ji, and Xiaoli Li. 2019. Heterogeneous graph attention networks for semi-supervised short text classification. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 4823–4832.
    Google ScholarLocate open access versionFindings
  • Amy Nguyen, Sasan Mosadeghi, and Christopher V Almario. 2017. Persistent digital divide in access to and use of the Internet as a resource for health information: Results from a California population-based study. International journal of medical informatics 103 (2017), 49–54.
    Google ScholarLocate open access versionFindings
  • Shivam B Parikh and Pradeep K Atrey. 2018. Media-rich fake news detection: A survey. In 2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR). IEEE, 436–441.
    Google ScholarLocate open access versionFindings
  • Fabian Pedregosa, Gaël Varoquaux, Alexandre Gramfort, Vincent Michel, Bertrand Thirion, Olivier Grisel, Mathieu Blondel, Peter Prettenhofer, Ron Weiss, Vincent Dubourg, et al. 2011. Scikit-learn: Machine learning in Python. Journal of machine learning research 12, Oct (2011), 2825–2830.
    Google ScholarLocate open access versionFindings
  • Martin Potthast, Johannes Kiesel, Kevin Reinartz, Janek Bevendorff, and Benno Stein. 2017. A stylometric inquiry into hyperpartisan and fake news. arXiv preprint arXiv:1702.05638 (2017).
    Findings
  • Hannah Rashkin, Eunsol Choi, Jin Yea Jang, Svitlana Volkova, and Yejin Choi. 2017. Truth of varying shades: Analyzing language in fake news and political fact-checking. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. 2931–2937.
    Google ScholarLocate open access versionFindings
  • Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme. 2009. BPR: Bayesian personalized ranking from implicit feedback. In Proceedings of the twenty-fifth conference on uncertainty in artificial intelligence. AUAI Press, 452–461.
    Google ScholarLocate open access versionFindings
  • Natali Ruchansky, Sungyong Seo, and Yan Liu. 2017. Csi: A hybrid deep model for fake news detection. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management. ACM, 797–806.
    Google ScholarLocate open access versionFindings
  • Daniel Scanfeld, Vanessa Scanfeld, and Elaine L Larson. 2010. Dissemination of health information through social networks: Twitter and antibiotics. American journal of infection control 38, 3 (2010), 182–188.
    Google ScholarLocate open access versionFindings
  • Franco Scarselli, Marco Gori, Ah Chung Tsoi, Markus Hagenbuchner, and Gabriele Monfardini. 2009. The Graph Neural Network Model. IEEE Transactions on Neural Networks 20, 1 (2009), 61–80.
    Google ScholarLocate open access versionFindings
  • Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. 2018. Modeling relational data with graph convolutional networks. In European Semantic Web Conference. Springer, 593–607.
    Google ScholarFindings
  • Baoxu Shi and Tim Weninger. 2016. Discriminative predicate path mining for fact checking in knowledge graphs. Knowledge-based systems 104 (2016), 123–133.
    Google ScholarLocate open access versionFindings
  • Kai Shu, Limeng Cui, Suhang Wang, Dongwon Lee, and Huan Liu. 2019. defend: Explainable fake news detection. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 395–405.
    Google ScholarLocate open access versionFindings
  • Jizhi Tang, Yansong Feng, and Dongyan Zhao. 2019. Learning to Update Knowledge Graphs by Reading News. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2632–2641.
    Google ScholarLocate open access versionFindings
  • Christopher C Tsai, SH Tsai, Q Zeng-Treitler, and BA Liang. 2007. Patientcentered consumer health social network websites: a pilot study of quality of user-generated health information. In AMIA Annu Symp Proc, Vol. 1137.
    Google ScholarLocate open access versionFindings
  • Sebastian Tschiatschek, Adish Singla, Manuel Gomez Rodriguez, Arpit Merchant, and Andreas Krause. 2018. Fake news detection in social networks via crowd signals. In Companion Proceedings of the The Web Conference 2018. 517–524.
    Google ScholarLocate open access versionFindings
  • Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. 2018. Graph Attention Networks. International Conference on Learning Representations (2018).
    Google ScholarLocate open access versionFindings
  • Soroush Vosoughi, Deb Roy, and Sinan Aral. 2018. The spread of true and false news online. Science 359, 6380 (2018), 1146–1151.
    Google ScholarLocate open access versionFindings
  • Yaqing Wang, Fenglong Ma, Zhiwei Jin, Ye Yuan, Guangxu Xun, Kishlay Jha, Lu Su, and Jing Gao. 2018. Eann: Event adversarial neural networks for multi-modal fake news detection. In Proceedings of the 24th acm sigkdd international conference on knowledge discovery & data mining. 849–857.
    Google ScholarLocate open access versionFindings
  • Jie Zhou, Ganqu Cui, Zhengyan Zhang, Cheng Yang, Zhiyuan Liu, Lifeng Wang, Changcheng Li, and Maosong Sun. 2018. Graph neural networks: A review of methods and applications. arXiv preprint arXiv:1812.08434 (2018). All the codes that we have implemented are available under the folder “Healthcare misinformation detection” through the following link: https://github.com/cuilimeng/DETERRENT.
    Findings
  • For the health-related article dataset, we manually created a dataset on healthcare by ourselves, under the folder “Dataset” at: https://github.com/cuilimeng/DETERRENT.
    Findings
Your rating :
0

 

Tags
Comments