Creating Precise Models by Discovering Long-term Dependencies in Process Trees

Given a log path and set of parameters, the dependency_miner algorithm is responsible for discovering long-term dependencies between the events and results into a precise Petri net by repairing the free-choice Petri net which includes the discovered rules. Added set of rules and computed evaluation metrics are returned.

Call miner(logpath, support, confidence, lift, soundness) It takes as input

    1. log_path (str): Path of event log
    2. support (str): Threshold value for support measure 
    3. confidence (str): Threshold value for confidence measure
    4. lift (str): Threshold value for lift measure, default min value = 1
    5. sound (str) : Soundness requirement if user wants sound model , "Yes/No"

The resulting precise Petri net can be found in the current location with the same name as that of input event log in .pnml and .svg format

Installation

pip install dependency_miner_pm4py

How to use it?

Install dependency_miner_pm4py package. Following, from dependency_miner.ltminer import miner

    Example: 
    log_path = "<path>\<file>.xes"
    support = "0.2"
    confidence = "0.3"
    lift = "1.0"
    sound = "Yes"
    miner(log_path, support, confidence, lift, sound)

License

Copyright (c) 2021 Ashwini Jogbhat

This repository is licensed under the MIT license. See LICENSE for details.