随着社交媒体的不断发展，学术知识的交流与分享行为发生了许多的变化，学术型用户生成内容越来越多的出现在各类社交媒体平台上。由于缺乏同行评审机制和进入社交媒体的无障碍性，每个人都可以在社交媒体中贡献许多学术内容，因此导致这些学术信息不仅数量庞大，而且质量参差不齐，这使得学者们很难找到高质量的信息来使用。本文报告了针对学术型用户生成内容的质量评估进行的相关研究，以期有助于识别高质量的学术信息。具体包括两个方面的研究内容。
      第一方面的研究为对针对一种类型的学术型用户生成内容进行了质量评估研究，即为对学术型社会化问答平台上问题的质量评估研究。此部分首先基于可以表征学术型问题质量的回复率，对影响学术型问题的回复率的因素进行了探索并进行了回复率的预测。结果发现问题的情感单词数量，面向未来的单词数量，常用动词单词数量，逗号数量，差异表达单词数量等语言特征可作为预测回复率的重要因素；接着，此部分提出结合回复率和提问者的权威度来衡量问题的质量，进而探索了影响问题质量的重要因素。结果发现学术型社会化问答平台上高质量的答案应具有较长的字数，较多的第二人称代词、助动词、朋友类词和标点符号中的问号，而不应该较多的具有影响力类词、第一人称单数代词和标点符号中的省略号。
      第二方面的研究是对学术任务信息检索情景下的学术型UGC质量评估的研究。本部分将对学术型用户生成内容质量评估扩充到了实际应用场景中，在更大范围内探索学者对检索出的各类学术型用户生成内容的质量评价情况，具体获取了评价学术型用户生成内容质量的标准，以及各个标准的重要性程度。最终构建了针对学术型用户生成内容的质量评估模型。

          With the continuous development of social media, many changes have taken place in the exchange and sharing of academic knowledge, and academic user-generated content has increasingly appeared on various social media platforms. Due to the lack of a peer review mechanism and accessibility to social media, everyone can contribute a lot of academic content to social media. As a result, this academic information is not only huge in quantity, but also of uneven quality, which makes it difficult for scholars to find high Quality information to use. This article reports on related research on the quality assessment of academic user-generated content, with a view to helping identify high-quality academic information. It specifically includes two aspects of research content.
          The first aspect of the research is to conduct quality evaluation research on one type of academic user-generated content, that is, to evaluate the quality of questions on academic social Q&A platforms. Based on the response rate that can characterize the quality of academic questions, this part explores the factors affecting the response rate of academic questions and predicts the response rate. The results found that the number of emotional words in the question, the number of future-oriented words, the number of common verb words, the number of commas, the number of differential expression words and so on can be used as important factors in predicting the response rate; then, this part proposes to combine the response rate and the authority of the questioner to measure the quality of the question, and then explore the important factors that affect the quality of the question. The results found that the high-quality answers on academic social Q&A platforms should have a longer word count, more second-person pronouns, auxiliary verbs, friend-like words, and question marks in punctuation, rather than more influential words, first-person singular pronouns and ellipsis.
         The second aspect of research is the quality evaluation of academic UGC in the context of academic task information retrieval. This section expands the quality evaluation of academic user-generated content to practical application scenarios, explores the quality evaluation of various academic user-generated content retrieved by scholars on a larger scale, and specifically obtains the quality evaluation criteria of academic user-generated content, and the importance of each criteria. Finally, a quality evaluation model for academic user-generated content is constructed.