DualKanbaFormer: An Efficient Selective Sparse Framework for Multimodal Aspect-based Sentiment Analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lawan, Adamu, Pu, Juhua, Yunusa, Haruna, Lawan, Muhammad, Umar, Aliyu, Yahya, Adamu Sani, Basi, Mahmoud
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916697171034112
author Lawan, Adamu
Pu, Juhua
Yunusa, Haruna
Lawan, Muhammad
Umar, Aliyu
Yahya, Adamu Sani
Basi, Mahmoud
author_facet Lawan, Adamu
Pu, Juhua
Yunusa, Haruna
Lawan, Muhammad
Umar, Aliyu
Yahya, Adamu Sani
Basi, Mahmoud
contents Multimodal Aspect-based Sentiment Analysis (MABSA) enhances sentiment detection by integrating textual data with complementary modalities, such as images, to provide a more refined and comprehensive understanding of sentiment. However, conventional attention mechanisms, despite notable benchmarks, are hindered by quadratic complexity, limiting their ability to fully capture global contextual dependencies and rich semantic information in both modalities. To address this limitation, we introduce DualKanbaFormer, a novel framework that leverages parallel Textual and Visual KanbaFormer modules for robust multimodal analysis. Our approach incorporates Aspect-Driven Sparse Attention (ADSA) to dynamically balance coarse-grained aggregation and fine-grained selection for aspect-focused precision, ensuring the preservation of both global context awareness and local precision in textual and visual representations. Additionally, we utilize the Selective State Space Model (Mamba) to capture extensive global semantic information across both modalities. Furthermore, We replace traditional feed-forward networks and normalization with Kolmogorov-Arnold Networks (KANs) and Dynamic Tanh (DyT) to enhance non-linear expressivity and inference stability. To facilitate the effective integration of textual and visual features, we design a multimodal gated fusion layer that dynamically optimizes inter-modality interactions, significantly enhancing the models efficacy in MABSA tasks. Comprehensive experiments on two publicly available datasets reveal that DualKanbaFormer consistently outperforms several state-of-the-art (SOTA) models.
format Preprint
id arxiv_https___arxiv_org_abs_2408_15379
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DualKanbaFormer: An Efficient Selective Sparse Framework for Multimodal Aspect-based Sentiment Analysis
Lawan, Adamu
Pu, Juhua
Yunusa, Haruna
Lawan, Muhammad
Umar, Aliyu
Yahya, Adamu Sani
Basi, Mahmoud
Computation and Language
Multimodal Aspect-based Sentiment Analysis (MABSA) enhances sentiment detection by integrating textual data with complementary modalities, such as images, to provide a more refined and comprehensive understanding of sentiment. However, conventional attention mechanisms, despite notable benchmarks, are hindered by quadratic complexity, limiting their ability to fully capture global contextual dependencies and rich semantic information in both modalities. To address this limitation, we introduce DualKanbaFormer, a novel framework that leverages parallel Textual and Visual KanbaFormer modules for robust multimodal analysis. Our approach incorporates Aspect-Driven Sparse Attention (ADSA) to dynamically balance coarse-grained aggregation and fine-grained selection for aspect-focused precision, ensuring the preservation of both global context awareness and local precision in textual and visual representations. Additionally, we utilize the Selective State Space Model (Mamba) to capture extensive global semantic information across both modalities. Furthermore, We replace traditional feed-forward networks and normalization with Kolmogorov-Arnold Networks (KANs) and Dynamic Tanh (DyT) to enhance non-linear expressivity and inference stability. To facilitate the effective integration of textual and visual features, we design a multimodal gated fusion layer that dynamically optimizes inter-modality interactions, significantly enhancing the models efficacy in MABSA tasks. Comprehensive experiments on two publicly available datasets reveal that DualKanbaFormer consistently outperforms several state-of-the-art (SOTA) models.
title DualKanbaFormer: An Efficient Selective Sparse Framework for Multimodal Aspect-based Sentiment Analysis
topic Computation and Language
url https://arxiv.org/abs/2408.15379