Personalized Hashtag Recommendation for Micro-videos

TL;DR

This paper introduces GCN-PHR, a graph convolution network model that effectively captures complex interactions among users, hashtags, and micro-videos for personalized hashtag recommendations, outperforming existing methods.

Contribution

The paper proposes a novel GCN-based model with attention mechanisms to model multi-type interactions for improved personalized hashtag suggestions.

Findings

Outperforms state-of-the-art methods by a large margin

Effectively models user, hashtag, and micro-video interactions

Uses attention to enhance message passing accuracy

Abstract

Personalized hashtag recommendation methods aim to suggest users hashtags to annotate, categorize, and describe their posts. The hashtags, that a user provides to a post (e.g., a micro-video), are the ones which in her mind can well describe the post content where she is interested in. It means that we should consider both users' preferences on the post contents and their personal understanding on the hashtags. Most existing methods rely on modeling either the interactions between hashtags and posts or the interactions between users and hashtags for hashtag recommendation. These methods have not well explored the complicated interactions among users, hashtags, and micro-videos. In this paper, towards the personalized micro-video hashtag recommendation, we propose a Graph Convolution Network based Personalized Hashtag Recommendation (GCN-PHR) model, which leverages recently advanced GCN techniques to model the complicate interactions among <users, hashtags, micro-videos> and learn their representations. In our model, the users, hashtags, and micro-videos are three types of nodes in a graph and they are linked based on their direct associations. In particular, the message-passing strategy is used to learn the representation of a node (e.g., user) by aggregating the message passed from the directly linked other types of nodes (e.g., hashtag and micro-video). Because a user is often only interested in certain parts of a micro-video and a hashtag is typically used to describe the part (of a micro-video) that the user is interested in, we leverage the attention mechanism to filter the message passed from micro-videos to users and hashtags, which can significantly improve the representation capability. Extensive experiments have been conducted on two real-world micro-video datasets and demonstrate that our model outperforms the state-of-the-art approaches by a large margin.