發展主題語料庫以輔助華語教學──以2019新型冠狀病毒語料庫為例

Developing a Topic-Specific Web Corpus, COVID-19, for Chinese Language Teaching and Learning

2019新型冠狀病毒（COVID-19）對人類產生巨大的影響，歐洲及美國皆有團隊建立英文COVID-19語料庫，但華語圈目前尚未有類似語料庫。因此，本文希望能補足此一缺口，建立「中文COVID-19主題語料庫」供研究人員、老師及學生使用。本研究的研究問題有二：（1）中文COVID-19語料庫和No Sketch Engine平臺能否提供有用的資訊？（2）此語料工具有何優缺點？本研究以WebBootCat技術建構COVID-19主題語料庫，也產出各種教學素材：（1）詞頻、（2）關鍵詞、（3）常見N連詞、（4）搭配詞。此研究發現WebBootCat技術可有效生成主題語料庫，此庫有以下優點：（1）即時性、（2）廣泛涵蓋率、（3）真實語言、（4）豐富語境。然而，此語料庫是爬取網路資料所建成，不免納入不相關的雜訊，而平臺上仍有許多重要工具有待開發。

The coronavirus disease 2019 (COVID-19) has had a serious impact on people around the globe. However, teams in Europe and the United States worked hard in developing English COVID-19 corpora. Yet, there is no such corpus available in Chinese. Therefore, this paper aimed to fill this gap by building a Chinese COVID-19 corpus for researchers, teachers, and students. The two research questions are as follows: (1) Can the Chinese COVID-19 corpus and No Sketch Engine platform provide useful information for Chinese teachers and students? (2) What are the advantages and disadvantages of this web platform in terms of the contents and analyses of the corpus? A Chinese COVID-19 corpus was built with WebBootCat. This study also generated raw data for assisting Chinese teaching and learning by analyzing the corpus: (1) top-frequency vocabulary items, (2) keywords, (3) n-grams, (4) collocations. This study found that using WebBootCat could efficiently generate a topic-specific corpus, which has the following advantages: (1) immediacy, (2) wide coverage, (3) authentic language, and (4) rich language contexts. However, as data were crawled from the web, irrelevant noises might be detected. Moreover, more tools need to be developed in No Sketch Engine.