DataAnalyser

This project use Scala (with sbt), Spark and MLlib to predict customers clicks on advertising.

Files structure

• artifacts: contains the compiled jars
  1. oracle.jar: the jar to predict new customers clicks
  2. model.jar: the jar to build the model
• common: contains the common code such as data cleaning stuff
• model: contains the code to build the model
• oracle: contains the code to predict new customers clicks

How to use it

This repo comes with a prebuilt model. You should have:

• Java 1.8;
• Scala 2.11.11;
• sbt 1.0.3;
• Spark 2.2.0.

1. Download or clone this repo
   git clone https://github.com/bern-ads/DataAnalyser
2. Go into the repo
   cd DataAnalyser
3. Go to the artifacts directory cd artifacts
4. Submit the spark job spark-submit --class "oracle.Main" --master local oracle.jar "<path_to_the_data_file>" "<path_to_selector_directory>" "<path_to_model_directory>"

You can find <path_to_selector_directory> at the root the repository. You can find <path_to_model_directory> at the root the repository. So the command should look like this: spark-submit --class "oracle.Main" --master local oracle.jar "<path_to_the_data_file>" "../bernads.spark.selector" "../bernads.spark.model"