Heterogeneous Network Representation Learning: A Unified Framework With Survey and Benchmark

Yang, Carl; Xiao, Yuxin; Zhang, Yu; Sun, Yizhou; Han, Jiawei

doi:10.1109/tkde.2020.3045924

Cited by 188 publications

(125 citation statements)

References 143 publications

Supporting

Mentioning

124

Contrasting

Unclassified

Order By: Relevance

“…We evaluated the use of our weighting scheme on just two representation learning methods: DeepWalk and TransE. Many other methods exist, especially for heterogeneous networks; a recent survey can be found in the study by Yang et al [ 27 ]. Despite the simplicity of these approaches, we obtained very good results, outperforming the state of the art for ADE prediction.…”

Section: Discussionmentioning

confidence: 99%

“…To obviate the need for manual extraction of features from text and graph inputs, representation learning aims to learn features or representations from the input directly, in an unsupervised manner. Representation learning from graphs is an active research area; see the studies by Goyal and Ferrara [ 26 ] and Yang et al [ 27 ] for general surveys and the study by Wang et al [ 28 ] for a survey on representation learning on KGs. The representations are vectorial representations of the vertices of the graph.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Adverse Drug Event Prediction Using Noisy Literature-Derived Knowledge Graphs: Algorithm Development and Validation

Dasgupta¹,

Jayagopal²,

Hong³

et al. 2021

JMIR Med Inform

View full text Add to dashboard Cite

Background Adverse drug events (ADEs) are unintended side effects of drugs that cause substantial clinical and economic burdens globally. Not all ADEs are discovered during clinical trials; therefore, postmarketing surveillance, called pharmacovigilance, is routinely conducted to find unknown ADEs. A wealth of information, which facilitates ADE discovery, lies in the growing body of biomedical literature. Knowledge graphs (KGs) encode information from the literature, where the vertices and the edges represent clinical concepts and their relations, respectively. The scale and unstructured form of the literature necessitates the use of natural language processing (NLP) to automatically create such KGs. Previous studies have demonstrated the utility of such literature-derived KGs in ADE prediction. Through unsupervised learning of the representations (features) of clinical concepts from the KG, which are used in machine learning models, state-of-the-art results for ADE prediction were obtained on benchmark data sets. Objective Due to the use of NLP to infer literature-derived KGs, there is noise in the form of false positive (erroneous) and false negative (absent) nodes and edges. Previous representation learning methods do not account for such inaccuracies in the graph. NLP algorithms can quantify the confidence in their inference of extracted concepts and relations from the literature. Our hypothesis, which motivates this work, is that by using such confidence scores during representation learning, the learned embeddings would yield better features for ADE prediction models. Methods We developed methods to use these confidence scores on two well-known representation learning methods—DeepWalk and Translating Embeddings for Modeling Multi-relational Data (TransE)—to develop their weighted versions: Weighted DeepWalk and Weighted TransE. These methods were used to learn representations from a large literature-derived KG, the Semantic MEDLINE Database, which contains more than 93 million clinical relations. They were compared with Embedding of Semantic Predications, which, to our knowledge, is the best reported representation learning method using the Semantic MEDLINE Database with state-of-the-art results for ADE prediction. Representations learned from different methods were used (separately) as features of drugs and diseases to build classification models for ADE prediction using benchmark data sets. The methods were compared rigorously over multiple cross-validation settings. Results The weighted versions we designed were able to learn representations that yielded more accurate predictive models than the corresponding unweighted versions of both DeepWalk and TransE, as well as Embedding of Semantic Predications, in our experiments. There were performance improvements of up to 5.75% in the F1-score and 8.4% in the area under the receiver operating characteristic curve value, thus advancing the state of the art in ADE prediction from literature-derived KGs. Conclusions Our classification models can be used to aid pharmacovigilance teams in detecting potentially new ADEs. Our experiments demonstrate the importance of modeling inaccuracies in the inferred KGs for representation learning.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Adverse Drug Event Prediction Using Noisy Literature-Derived Knowledge Graphs: Algorithm Development and Validation

Dasgupta¹,

Jayagopal²,

Hong³

et al. 2021

JMIR Med Inform

View full text Add to dashboard Cite

show abstract

“…HINs (Heterogeneous Information Network) have been proposed to model complex objects and their rich relations. HIN embedding aims to embed the nodes in the network into a low-dimensional space, which help to facilitate downstream applications such as link prediction, personalized recommendation, node classification, etc [22,30]. One line of work leverages meta-path-based contexts for semantic-preserving embedding.…”

Section: Heterogeneous Information Network Embeddingmentioning

confidence: 99%

Multi-view Denoising Graph Auto-Encoders on Heterogeneous Information Networks for Cold-start Recommendation

Zheng

et al. 2021

Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

Cold-start recommendation is a challenging problem due to the lack of user-item interactions. Recently, heterogeneous information network (HIN)-based recommendation methods use rich auxiliary information to enhance users and items' connections, helping alleviate the cold-start problem. Despite progress, most existing methods model HINs under traditional supervised learning settings, ignoring the gaps between training and inference procedures in cold-start scenarios. In this paper, we regard cold-start recommendation as a missing data problem where some user-item interaction data are missing. Inspired by denoising auto-encoders that train a model to reconstruct the input from its corrupted version, we propose a novel model called Multi-view Denoising Graph Auto-Encoders (MvDGAE) on HINS. Specifically, we first extract multifaceted meaningful semantics on HINs as multi-views for both users and items, effectively enhancing user/item relationships on different aspects. Then we conduct the training procedure by randomly dropping out some user-item interactions in the encoder while forcing the decoder to use these limited views to recover the full views, including the missing ones. In this way, the complementary representations for both users and items are more informative and robust to adjust to cold-start scenarios. Moreover, the decoder's reconstruction goals are multi-view user-user and item-item relationship graphs rather than the original input graphs, which make the features of similar users (or items) in the meta-paths closer together. Finally, we adopt a Bayesian task weight learner to balance multi-view graph reconstruction objectives automatically. Extensive experiments on both public benchmark datasets and a large-scale industry dataset WeChat Channel demonstrate that MvDGAE significantly outperforms the state-of-the-art recommendation models in various cold-start scenarios. The case studies also illustrate that MvDGAE has potentially good interpretability. CCS CONCEPTS• Information systems → Data mining; Collaborative filtering.

show abstract

“…With the rapid development of web 2.0 and mobile networks, event detection on heterogeneous data has drawn more attention in recent years. Yang et al [38] proposed a Complexity unified model to dynamically learn how to represent the data with different features of a heterogeneous social network. Liu et al [39] treated the breaking news as a heterogeneous social data stream and developed how to extract events from the dynamic data stream.…”

Section: Dynamic Community Detection Of Heterogeneous Socialmentioning

confidence: 99%

A Novel Emerging Topic Identification and Evolution Discovery Method on Time‐Evolving and Heterogeneous Online Social Networks

Zhang

et al. 2021

Complexity

View full text Add to dashboard Cite

With the fast development of web 2.0, information generation and propagation among online users become deeply interweaved. How to effectively and immediately discover the new emerging topic and further how to uncover its evolution law are still wide open and urgently needed by both research and practical fields. This paper proposed a novel early emerging topic detection and its evolution law identification framework based on dynamic community detection method on time-evolving and scalable heterogeneous social networks. The framework is composed of three major steps. Firstly, a time-evolving and scalable complex network denoted as KeyGraph is built up by deeply analyzing the text features of all kinds of data crawled from heterogeneous online social network platforms; secondly, a novel dynamic community detection method is proposed by which the new emerging topic is detected on the modeled time-evolving and scalable KeyGraph network; thirdly, a unified directional topic propagation network modeled by a great number of short texts including microblogs and news titles is set up, and the topic evolution law of the previously detected early emerging topic is identified by fully utilizing local network variations and modularity optimization of the “time-evolving” and directional topic propagation network. Our method is proved to yield preferable results on both a huge amount of computer-generated test data and a great amount of real online network data crawled from mainstream heterogeneous social networks.

show abstract

Heterogeneous Network Representation Learning: A Unified Framework With Survey and Benchmark

Cited by 188 publications

References 143 publications

Adverse Drug Event Prediction Using Noisy Literature-Derived Knowledge Graphs: Algorithm Development and Validation

Adverse Drug Event Prediction Using Noisy Literature-Derived Knowledge Graphs: Algorithm Development and Validation

Multi-view Denoising Graph Auto-Encoders on Heterogeneous Information Networks for Cold-start Recommendation

A Novel Emerging Topic Identification and Evolution Discovery Method on Time‐Evolving and Heterogeneous Online Social Networks

Contact Info

Product

Resources

About