Loading...

Representation Learning for Heterogeneous Information Networks

Mirzaie, Mohammad Ali | 2021

779 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54300 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Beigy, Hamid
  7. Abstract:
  8. Around world and the networks within it can be modeled in various templates. Graph structure is one of those templates in which objects and relations may have more than one types. We call this phenomenon "heterogeneity".Heterogeneity makes the networks hard to model and that is why the proposed methods for modeling the networks assumed the network structures homogeneous. This assumption may cause data loss due to ignoring the variety of types in network objects and relations and this loss can lessen the accuracy of data mining tasks.To tackle the challenge of data loss in the mentioned assumption, learning representations for heterogeneous information networks (HINs) was introduced. HINs consider the variety of types between network objects and also distinguish between multiple types of relations between network objects.Despite of old methods which consider heterogeneous graphs as homogeneous ones which may cause challenges in extracting the deep semantics from the information network, new methods consider the heterogeneity of information networks in their models.In this thesis we propose a novel Deep Structural Semantics Learning (DSSL) model for HINs embedding. Our model uses the attention mechanism in a hierarchical manner in order to determine the best way to embed graph objects and also we employ smart aggregation units to extract structural features of objects from their most effective neighbors. Experiments conducted on two open-world public datasets show the superiority of our proposed model over previous state-of-the-art solutions
  9. Keywords:
  10. Graph Neural Network ; Deep Learning ; Attention Mechanism ; Representation Learning ; Machine Learning ; Heterogeneous Information Networks (HINs)

 Digital Object List

 Bookmark

...see more