Multimodal Sentiment Analysis (MSA) utilizes multimodal data to infer the users' sentiment. Previous methods focus on equally treating the contribution of each modality or statically using text as the dominant modality to conduct interaction, which neglects the situation where each modality may become dominant. In this paper, we propose a Knowledge-Guided Dynamic Modality Attention Fusion Framework (KuDA) for multimodal sentiment analysis. KuDA uses sentiment knowledge to guide the model dynamically selecting the dominant modality and adjusting the contributions of each modality. In addition, with the obtained multimodal representation, the model can further highlight the contribution of dominant modality through the correlation evaluation loss. Extensive experiments on four MSA benchmark datasets indicate that KuDA achieves state-of-the-art performance and is able to adapt to different scenarios of dominant modality.

Kuda is a fintech company on a mission to make financial services accessible, affordable and rewarding for every African on the planet. We're a tribe of passionate and diverse people who dreamed of building an inclusive money app that Africans would love so it's only right that we ended up with the name'Kuda' which means'love' in Shona, a language spoken in the southern part of Africa. We're giving Africans around the world a better alternative to traditional finance by delivering free money transfers, smart budgeting and instant access to credit through digital devices. We've raised over $90 million from some of the world's most respected institutional investors, and we're rolling out our game-changing services globally from our offices in Nigeria, South Africa, and the UK. To support our rapid growth and reach 4M customers by the end of 2022, Kuda is looking for an experienced Data Analyst who can mine attribution, engagement and behavioral data to establish insights and ensure we make data-driven decisions around product and channel experimentations, ensuring we focus on the highest ROI growth tactics.

TOKYO--The 1980s movie "Back to the Future Part II" transports viewers to a 2015 where people pay for cab rides with the swipe of a finger. In the real 2016, fingerprints have entered the mainstream as a way to log on to smart devices, but it's still difficult to pay for rides or at the cash register with just a finger. To be successful, a payment system would have to match the print against a database that may contain thousands or millions of other prints, and--to keep impatient customers happy--do it within seconds. That's the challenge Yasuhiro Kuda says his Tokyo-based startup, Liquid Inc., has addressed. Backed by the Japanese government and some of the nation's biggest names in finance and technology--including the venture-capital arms of banking giant Mizuho Financial Group Inc. MFG -0.31 % and cellphone provider NTT DoCoMo Inc.--Liquid is rolling out a fingerprint-payment system at retailers in Japan and elsewhere in Asia, including Sri Lanka and the Philippines.