Publication Type

Journal Article

Version

acceptedVersion

Publication Date

7-2021

Abstract

Multimodal recommender systems alleviate the sparsity of historical user-item interactions. They are commonly catalogued based on the type of auxiliary data (modality) they leverage, such as preference data plus user-network (social), user/item texts (textual), or item images (visual) respectively. One consequence of this categorization is the tendency for virtual walls to arise between modalities. For instance, a study involving images would compare to only baselines ostensibly designed for images. However, a closer look at existing models' statistical assumptions about any one modality would reveal that many could work just as well with other modalities. Therefore, we pursue a systematic investigation into several research questions: which modality one should rely on, whether a model designed for one modality may work with another, which model to use for a given modality. We conduct cross-modality and cross-model comparisons and analyses, yielding insightful results pointing to interesting future research directions for multimodal recommender systems.

Keywords

Data Models, Visualization, Recommender Systems, Analytical Models, Predictive Models, Computational Modeling, Matrices, Multimodal Recommender Systems, Multimodality, Cross Modality

Discipline

Databases and Information Systems | Data Science

Research Areas

Data Science and Engineering

Publication

IEEE Internet Computing

Volume

25

Issue

4

First Page

50

Last Page

57

ISSN

1089-7801

Identifier

10.1109/MIC.2021.3059027

Publisher

IEEE Computer Society

Embargo Period

5-20-2021

Copyright Owner and License

Authors

Additional URL

https://doi.ieeecomputersociety.org/10.1109/MIC.2021.3059027

Download

Share

COinS