本文对学习者语料库在汉语中介语混淆词语研究中的效用进行了多角度评估。统计数据显示,语料库对原始语料信息的保留和标注的准确度使统计结果具有较高程度的可信度,但语料库的规模、话题数量等有限,学习者母语背景及汉语水平等级分布也不够均衡,语料库偏误标注还存在遗漏和偏差,这些因素使统计结果具有一定的局限性。 This paper evaluates the utility of learner corpus in the study of Chinese interlanguage confusion words. Statistics show that the corpus retains the original corpus information and the accuracy of annotation makes the statistical results have a high degree of credibility, but the size of the corpus, the number of topics is limited, the learner background and the level of Chinese language level distribution is not balanced There are still omissions and deviations in the corpus error annotation. These factors make the statistical results have some limitations.