Datas of Disease Patterns

1.the "dingxiang_datas.xls"contains all the original data which is crawling from DingXiang forum, and also the word segmentation result for each medical record in the dataset.<br>2.the "pmi_new_words.txt" is the result of new medical word found by calculating mutual information, and the "new_medical_dict.txt" is the updated dictionary which we have used for word segmentation.<br>3.the "structure_medical_record.txt" is the structured text after the word segmentation for original data, which contains the symptoms and diseases in each medical record，and the symptom word frequency is showed in "wordfreq.txt".<br>4.the"0.3_disease-disease_rules.txt"、"0.3_symptom-disease_rules.txt"、"0.3_symptom-symptom_rules.txt"are association rules mined from the dataset which h-confidence threshold is set 0.3 and support threshold is set 0.0001.<br>5.the "network_data.txt" contains the data for building the symptom-disease bipartite network.<br>6.the "network_communities.csv" describes the communities in which each disease belongs.<br>p.s. if you encounter a "d", it means the word is a disease description vocabulary, and "z" or "s" represents a symptom description vocabulary.