SPL File PredictiveBlocklisting.spl

Content

Operators
  • PredictiveBlocklisting: The PredictiveBlocklisting operator analyzes the feature vector produced by the PredictiveBlocklistingFE operator using an SPSS model.

Composites

composite PredictiveBlocklisting(output PredictiveBlocklistingResults; input PredictiveBlocklistingInput, SPSSModelFileNames)

SPL composite operator image not displayed. Problem loading file: ../../image/tk$com.teracloud.streams.cybersecurity/op$com.teracloud.streams.cybersecurity.analytics$PredictiveBlocklisting.svg

The PredictiveBlocklisting operator analyzes the feature vector produced by the PredictiveBlocklistingFE operator using an SPSS model. The model predicts whether a domain should be blocklisted. For each feature vector analyzed, the operator submits a tuple predicting whether the domain is "supicious" or "benign".

Input Ports This operator contains 2 input ports.

Parameters

  • pimfile: Specifies the absolute path of the SPSS pim file.
  • parfile: Specifies the absolute path of the SPSS par file.
  • xmlfile: Specifies the absolute path of the SPSS xml file.

Input Ports

  • PredictiveBlocklistingInput: Input port 0 ingests tuples containing the feature vector produced by the PredictiveBlocklistingFE operator. For convenience, the com.teracloud.streams.cybersecurity.types::PredictiveBlocklisting_t type can be used as the input tuple type for this operator.
  • SPSSModelFileNames: Input port 1 is a control port that allows for dynamically updating the SPSS model. This port ingests a tuple containing an attribute of type "rstring". Typically this port is connected to an upstream DirectoryScan that scans a directory for updated SPSS model files. Output Ports This operator contains 1 output port.

Output Ports

  • PredictiveBlocklistingResults: Output port 0 submits a tuple for each feature vector analyzed. This operator expects that the output tuple type be the same as the input tuple type, and that this type contains two rstring attributes called "predictedClass" and "confidence. The operator will populate the "predictedClass" and "confidence" attributes with the results of the SPSS model. Therefore, the com.teracloud.streams.cybersecurity.types::PredictiveBlocklisting_t type can be used as the output tuple type.