hyperdoc2vec: Distributed Representations of Hypertext Documents

TL;DR

Hyperdoc2vec is a novel embedding method designed for hypertext documents that preserves essential information and outperforms existing models in tasks like paper classification and citation recommendation.

Contribution

The paper introduces hyperdoc2vec, a new hyper-document embedding approach with four criteria, and demonstrates its superiority over existing methods through systematic evaluation.

Findings

Hyperdoc2vec outperforms competitors in paper classification.

Hyperdoc2vec improves citation recommendation accuracy.

The model effectively preserves critical hyper-document information.

Abstract

Hypertext documents, such as web pages and academic papers, are of great importance in delivering information in our daily life. Although being effective on plain documents, conventional text embedding methods suffer from information loss if directly adapted to hyper-documents. In this paper, we propose a general embedding approach for hyper-documents, namely, hyperdoc2vec, along with four criteria characterizing necessary information that hyper-document embedding models should preserve. Systematic comparisons are conducted between hyperdoc2vec and several competitors on two tasks, i.e., paper classification and citation recommendation, in the academic paper domain. Analyses and experiments both validate the superiority of hyperdoc2vec to other models w.r.t. the four criteria.