# Multi-Modal Adversarial Autoencoders for Recommendations of Citations and Subject Labels

## TL;DR

This paper introduces multi-modal adversarial autoencoders for citation and subject label recommendation, demonstrating that adversarial regularization improves performance and that the semantics of item co-occurrence influence model effectiveness.

## Contribution

The study systematically evaluates multi-modal adversarial autoencoders across two recommendation tasks, highlighting the importance of co-occurrence semantics and input modalities.

## Key findings

- Adversarial regularization consistently enhances recommendation performance.
- Item co-occurrence semantics differ between citation and subject label tasks.
- Partial item set input is beneficial only when co-occurrence indicates relatedness.

## Abstract

We present multi-modal adversarial autoencoders for recommendation and evaluate them on two different tasks: citation recommendation and subject label recommendation. We analyze the effects of adversarial regularization, sparsity, and different input modalities. By conducting 408 experiments, we show that adversarial regularization consistently improves the performance of autoencoders for recommendation. We demonstrate, however, that the two tasks differ in the semantics of item co-occurrence in the sense that item co-occurrence resembles relatedness in case of citations, yet implies diversity in case of subject labels. Our results reveal that supplying the partial item set as input is only helpful, when item co-occurrence resembles relatedness. When facing a new recommendation task it is therefore crucial to consider the semantics of item co-occurrence for the choice of an appropriate model.