config.yaml

sparkSettings:
  appName: "Twitter-Opinion-Mining"
  executorMemory: "10g"
  driverMemory: "8g"
input:
  tweets: "wasb://ms-tweets-medium@aaltospark.blob.core.windows.net/"
  training: "wasb://sentiment-training@aaltospark.blob.core.windows.net/twitter_training.csv"
  stopwords: "wasb://sentiment-stopwords@aaltospark.blob.core.windows.net/stopwords.txt"
output:
  csvOutputPath: "wasb://sentiment-output@aaltospark.blob.core.windows.net/output.csv"
  hostname: "aaltoms"
  database: "ms_twitter"
  user: "aaltoadmin@aaltoms"
  password: "fD^bS7yVGw"
  encrypt: true
  trustServerCertificate: false
  hostNameInCertificate: "*.database.windows.net"
  loginTimeout: 30
  table: "twitter_opinions"
topicModelling:
  i: 20             # The number of iterations performed by the topic modelling
  k: 3              # The number of topics to be generated by the topic modelling
  n: 10             # The number of words generated per topic
  wnd: 5            # The topic modelling window, i.e. the number of time intervals that are to be considered part of a peak. E.g. if there is a peak at t=10, and tmWnd is 3, tweets will be analyzed from t=9 to (and including) t=11. NB! Must be an odd, positive number.
peakDetection:
  #primaryFilter:
  #secondaryFilter:
  alpha: 5          # The roughness of the primary peak detection filter (defines moving average window size)
  beta: 12          # The roughness of the secondary peak detection filter (defines moving average window size). By definition, beta > alpha.
  #delta_gen: