I-KAHAN: Image-Enhanced Knowledge-Aware Hierarchical Attention Network for Multi-modal Fake News Detection Øystein L. Nilsen 1 , Pelin Mi¸ se 2 , Ahmet Yıldız 2 , Eniafe Festus Ayetiran 3,4[0000-0002-6816-2781] , and ¨ Ozlem ¨ Ozg¨ obek 3[0000-0003-2612-2009] 1 Sopra Steria, Oslo, Norway oystein.lnilsen@soprasteria.com 2 MEF University, Department of Computer Engineering, Istanbul, Turkey {misepe, yildizah}@mef.edu.tr 3 Norwegian University of Science and Technology (NTNU), Department of Computer Science, Trondheim, Norway {ozlem.ozgobek, eniafe.ayetiran}@ntnu.no 4 Achievers University, Department of Computer Science, Owo, Nigeria eniafe.ayetiran@achievers.edu.ng Abstract. In the quest to combat the proliferation of fake news, ac- curate detection of fabricated news content has become increasingly desirable. While existing methodologies leverage a variety of news at- tributes, such as text content and social media comments, few incorpo- rate diverse features from different modalities like images. In this pa- per, Image-Enhanced Knowledge-Aware Hierarchical Attention Network (I-KAHAN) architecture is proposed as an enhancement to the existing KAHAN architecture. The I-KAHAN architecture utilizes a wide variety of attributes including news content, user comments, external knowledge, and temporal information which are inherited from the KAHAN archi- tecture, and extends it by integrating image-based information as an additional feature. This work contributes to refining and expanding fake news detection methodologies by embracing a more comprehensive range of features and modalities, and offers valuable insights into the effective- ness of various methods for the numerical representation of images, fea- ture aggregation and dimensionality reduction. Experiments conducted on two real-world datasets, PolitiFact and GossipCop, assessing the per- formance of the I-KAHAN architecture, demonstrated approximately 3% improvement in accuracy over the KAHAN architecture, highlighting the potential benefits of incorporating diverse features and modalities for en- hanced fake news detection performance. Keywords: Fake News Detection · Deep Learning · Multi-Modality. 1 Introduction Rapid advancements in digital technology gives people easier access to news via websites or social media. Even though this may be advantageous, it also brings