Applying Text and Image Mining Techniques for Identifying Future Signals of AI-Powered Products

Since the debut of Sora earlier this year, AI products have rapidly integrated into everyday life, with Apple's launch of the Apple Intelligence AI phone being a prime example. Exploring the impact of AI on smart products helps accelerate their integration, offering smarter and more convenient services. This study proposes a multimodal approach to detect future signals of AI-powered products, utilizing the extended BERTopic model to integrate text and image data, and employing theme filtering and meaning construction methods to extract and expand weak signals. By collecting a multimodal dataset from Sina News and validating it with the Apple AI phone case, the study demonstrates the effectiveness of future signals, providing valuable guidance for the development of AI-powered products and expanding the application of future signals in multimodal data analysis.