An Empirical Analysis of Factors Influencing Customer Loyalty in Sichuan's Online Auto Sales in China - Mark Ma Lang

Driven by the rapid advancement of artificial intelligence (AI), particularly breakthroughs in deep learning (DL) and large language models (LLMs), corpus linguistics is undergoing a paradigm shift from data-driven to intelligence-driven research. This article systematically reviews the current state and prevailing challenges of AI-empowered corpus linguistics. Our analysis identifies four major research themes: intelligent and automated corpus processing, deepening and innovation in corpus analysis paradigms, the rise and integration of multimodal corpora, and emerging ethical and methodological concerns. Meanwhile, Key problems include inadequate corpus quality control (compromising authenticity, representativeness, and standardization), limited model interpretability (the “black box” dilemma that constrains theoretical contributions), and insufficient interdisciplinary integration due to disconnection between linguistic theory and technical applications.

Finally, the article outlines future directions, arguing that research should focus on human-machine collaborative annotation paradigms, the development of dynamic and embodied corpora, exploration of interpretability and causal inference, the establishment of ethical governance frameworks, and deeper interdisciplinary engagement. These efforts aim to provide theoretical and methodological support for the reconstruction and innovation of corpus linguistics in the AI age.