近年來，有很多的文獻研究顯示，異常的miRNA表現量在發現各種不同的癌症上扮演著重要的角色，為了辨別miRNA是如何影響癌症的成長或者抑制，了解miRNA異常的表現量是如何影響癌症是必要的。然而，只尋找miRNA在辨認miRNA與癌症的關係上是不夠的，我們需要在生物醫學的文章中進一步的尋找癌症狀態改變的訊息。為了有效率的存取，miRNA表現量與癌症狀態改的關係，我們運用了不同的文字探勘工具，去分析在文獻中所提及的miRNA表現量與癌症狀態改變。很可惜的是，目前既有的miRNA與癌症關聯的資料庫缺乏癌症狀態改變與miRNA表現量關係的資料，而且在miRNA表現量與癌症狀態改變的關係上的查詢是不足的。本論文提出了一個miRNA表現量跟癌症狀態改變關聯的文字探勘網路平台並且為使用者建立了更完整的查詢介面，miRCSC，包含了miRNA表現量影響癌症狀態變化的文獻，而且我們將公開我們的手動標記的資料集。我們期望我們的搜尋平台在改善治療與偵測癌症上有所幫助。;Currently, lots of researches have shown aberrant miRNA expression event plays important role in many aspects of tumor development. In order to identify miRNAs that are involved in tumor progression or tumor regression, it is essential to understand how such abnormal expression affect cancers. However, identify miRNA expression event alone is not enough for miRNA-cancer relationship, we need further info of cancer state change in such biomedical literature. In our work, we define new query concept, cancer state change; which existing web tools lack of. For efficient access to the result of such relationship in biomedical literature, we apply different text mining tools of miRNA expression event extracted and cancer state change identified in biomedical literatures. Unfortunately, existing databases lack of cancer state change info and also not strength enough to query that specify how cancer state change along with the change in miRNA expression event. In this paper, we present a web platform about miRNA expression event and cancer state change relationships based on text mining approaches and create more comprehensive query interface for users. miRCSC contains articles about miRNA expression event involved in cancer state change and also we released our manual annotated data. We’re expecting our released tool helps improving of diagnosis and treatment of human cancers.
