摘要

外语教学评价存在"评价即测量"的固有观念,评价内容无法完全对标课程目标,且现有评价模式无法使学生获得及时有效的反馈.生成式人工智能的迅速发展为外语教学评价的根本性变革带来重要契机,能够大幅提升教学评价的效率、创新评价模式、提高评价的个性化水平.然而,人工智能在赋能外语教学评价的同时也存在一定局限性,这主要体现在:生成内容的准确性和客观性有待验证、"智能生成"造成教学评价不公、"智能评价"导致评价主体不明.因此,应通过转变评价理念、智能生成高质量测评资源、构建多元教学评价模式、严格审核评价各个环节、强化学术诚信理念,规避人工智能赋能外语教学评价的局限性,使人工智能有效赋能外语教学评价.; Foreign language teaching assessment has long been constrained by the entrenched notion that"assessment equals measurement."As a result,assessment content often fails to fully align with curricular objectives,and existing assessment mod-els are unable to provide students with timely and effective feedback.The rapid development of generative artificial intelligence(GAI)presents a significant opportunity for the fundamental transformation in foreign language teaching assessment,and GAI can substantially enhance the efficiency for the,innovate its models,and enhance its personalization.However,while empower-ing foreign language teaching assessment,GAI also exhibits certain limitations.These are mainly manifested in the unconfirmed accuracy and objectivity of generated content,the potential for"intelligent generation"to lead to inequities in assessment out-comes,and the ambiguity of assessment agency resulting from"intelligent evaluation."Accordingly,it is necessary to mitigate these limitations by transforming assessment philosophies,leveraging intelligent technologies to generate high-quality assessment resources,constructing diversified assessment models,rigorously reviewing each stage of the evaluation process,and strengthen-ing the principle of academic integrity.Through these measures,GAI can be effectively integrated into foreign language teaching assessment to optimize its overall effectiveness.

关键词