ADRs_Extract(f不良反应抽取)

To see how the system work, just look at the Method.bmp ,that's a workflow of our system.

INPUT　DATA

• In the input/ folder, we need an articles.csv file to save the users' text what you collected from the web. It's content like this:

  100001, Good morning Deepdive. I hate you.
  100002, Oh you break my heart!
  ...
  

  The first part is the ID of articles, follow by a comma.

  The Second part is the content of article which should not have any comma (because csv use comma to separate the different part in .csv )

• In the input/ folder, we need an ards_dbdata.csv file to save the known ADRs of specific Medicine. It's content like this:

  优甲乐,头晕
  优甲乐,胃痛
  ...
  

  The first part is the Name of target Medicine. The second part is the ARDs of that Medicine.

• In the udf/reco_med_ard_guo.py, the function need a list parameter which contain every ADR of all Medicine we know. It is like this:

  list ards=['血压升高','头晕',........]  #in python language