Tag Recommendation by Word-Level Tag Sequence Modeling

TL;DR

This paper presents a novel word-level sequence-to-sequence model for tag recommendation, leveraging LSTM and attention mechanisms to improve performance over existing text classification methods.

Contribution

It introduces a new approach transforming tag recommendation into a text generation task using sequence-to-sequence modeling with attention.

Findings

Outperforms state-of-the-art text classification methods on Zhihu datasets

Utilizes LSTM encoder and attention-based decoder for effective tag sequence modeling

Demonstrates the effectiveness of word-level modeling in tag recommendation

Abstract

In this paper, we transform tag recommendation into a word-based text generation problem and introduce a sequence-to-sequence model. The model inherits the advantages of LSTM-based encoder for sequential modeling and attention-based decoder with local positional encodings for learning relations globally. Experimental results on Zhihu datasets illustrate the proposed model outperforms other state-of-the-art text classification based methods.