# Introduction The economic and financial world is influenced by many complex factors, such as human decisions, emotions, and collective psychology. However, time series forecasting models are often limited by the use of external data, such as news and social media. # How HINTS works The new study proposes a machine learning framework called HINTS (Human Insights Through Networked Time Series). This model uses the Friedkin-Johnsen (FJ) opinion dynamics model as a structural inductive bias to model evolving social influence, memory, and bias patterns. # Characteristics of the model The HINTS model extracts human factors from time series residuals without using external data. This approach improves forecasting accuracy and allows for more detailed analysis of social dynamics. # Experimental results The study's results show that HINTS improves forecasting accuracy compared to traditional models. Additionally, case studies and ability analyses have demonstrated the interpretability of HINTS, showing a strong correlation between the extracted factors and real-world events. # Conclusion The new study presents an important step forward in understanding social dynamics and time series forecasting. The HINTS approach offers a new perspective for analyzing economic data and financial information.