对 RDD 的每个实例进行分类 | Apache 火花 Scala

Question

I'm starting to work with RDD's and I have some doubts. In my case, I have a RDD and I want to classify his data. My RDD contains the following:

Array[(String, String)] = Array((data: BD=bd_users,BD_classified,contains_people, rbd: BD=bd_users,BD_classified,contains_people),
(data: BD=bd_users,BD_classified,contains_people,contains_users, user: id=8282bd, BD_USERS,bdd),
(data: BD=bd_experts,BD_exp,contains_exp,contains_adm, rbd: BD=bd_experts,BD_ea,contains_exp,contains_adm),
(data: BD=bd_test,BD_test,contains_acc,contains_tst, rbd: BD=bd_test,BD_test,contains_tst,contains_t))

As you can see the RDD contains two strings, the first one start with data and the second one starts with rbd. What I want to do is classify every instance of this RDD as you can see here:

If the instance contains bd_users & BD_classified -> users
bd_experts & BD_exp -> experts
BD_test -> tests

The output would be something like this for this RDD:

1. Users
2. Users
3. Experts
4. Test

To do this I would like to use a map that calls a function for every instance in this RDD but I don't know how can orientate this:

val rdd_groups = rdd_1.map(x=>x(0).toString).map(x => getGroups(x))
def getGroups(input: String): (String) = {
//here i should use for example case to classify this strings?
}

If you need something more or examples, just tell me it. Thanks in advance!

Answer 1

Well assuming you have a RDD of strings and a classifier already defined:

  val rdd: RDD[String] =
    ???

  def classify(input: String): String =
    ???
  
  rdd.map(input => classify(input))