Application of Semantic Computing in Cancer on Secondary Data Analysis

There have been an enormous number of publications on cancer research. These integrated but unstructured cancer-related articles are of great value for cancer diagnosis, treatment and prevention. In this study, we use the basic concepts underlying text mining and semantic computing to discuss the current state-of-the-art text mining applications in cancer research. Using the abstract of literature extracted from PubMed between 2008 and 2016, a total 925,648 articles are used for subsequent text mining. Among the 925,648 articles, the top 5 most studied cancer types were breast cancer (23.82%), lung cancer (10.54%), prostate cancer (9.90%), rectal cancer (8.44%), and ovarian cancer (4.44%). The top 3 most frequently occurred keywords in the abstracts of the 925,648 articles are patients, cancer, and cell where each appear 1,445,688, 1,284,140, and 676,924 times, respectively. Analysis of the key concepts indicate that the most common concepts are patients, cancer, cell and tumor. Our results suggest that while the risk factors of cancer, treatment of cancer, and survival of cancer patients were popular research topics, end-of-life cancer care issues are less studied. Further studies should explore these areas since they are as important as treatment of the disease itself for many patients.