Large language model-driven multi-topic and multi-level sentiment analysis framework for futures market news

To address the need for the systematic and in-depth mining of complex sentiment signals from futures market news, this paper proposes a large language model -driven multi-topic and multi-level sentiment analysis framework. Through a "macro-topic-specific aspect/event" hierarchical strategy, the framework constructs a topic system that covers multi-dimensional market elements and achieves accurate discrimination of topic-level sentiments. By incorporating aspect-sentiment-opinion triplet extraction and event sentiment analysis techniques, it identifies the sentiment impact of key market elements and unexpected events. Low-rank adaptation is employed to achieve efficient domain adaptation of a localized large language model, validating the applicability of large language models in financial text analysis. Experimental results show that the proposed framework performs well across multiple sentiment analysis tasks, effectively distinguishing the sentiment tendencies of different topics and deeply mining complex sentiment signals within news texts. This research provides a systematic solution for in-depth sentiment analysis of futures market news and offers reliable technical support for the quantitative analysis of the impact of news sentiment on the futures market.