In the paper Transforming Object-Centric Event Logs to Temporal Event Knowledge Graphs we introduce and formalize the concept of Temporal Event Knowledge Graphs (tEKGs), which are designed to record object-centric event data. This approach facilitates the tracking of changes in entity attributes over time. We implemented the transformation tool in Python and used three publicly available log files presented in OCEL 2.0 to transform into the tEKG format. We developed the transformation tool in three modes: batch import, live import, and hybrid mode. All three modes of the transformation tool are compared in terms of time taken for conducting the transformation. This repository provides material for this paper, including the source code for transforming OCEL 2.0 to tEKG, a running example test file, and experiment results. The final transformed datasets can be found at Zenodo (anonymized).
In this paper, we used three OCEL 2.0 data sets, which are all publicly available. In the links below, you can download each of theese files.
1. Container Logistics Object-centric Event Log: This is an artificial event log according to the OCEL 2.0 Standard, simulated using CPN-Tools. 2. Order Management Object-centric Event Log: This is a simulation of an extension of the order management log in the former OCEL standard.
3. Procure-To-Payment (P2P) Object-centric Event Log: This simulation extensively uses genuine SAP transactions and object types to offer a realistic representation of the P2P process.
Once you have download all the datasets, you can place them in the directory ocel2. Additionally, we have provided an OCEL 2.0 file as a running example that we used in the paper, named runningExample-course.jsonocel.
The transformation tool is developed in three modes of batch, live, abd hybrid in Python 3.11.8. To use this tool, you need to take the following steps:
1. Make sure you have installed Python on your computer.
2. install all the packages, which are available in requirements.txt. You can install all the packages using the following command:
pip install -r requirements.txt3. Once you installed all the packages, lunch the jupyter lab using the following command:
jupyter labOnce you lunch the jupyter lab, you can see there are three python notebooks, namely: import batch.ipynb, live import.ipynb, and hybrid import.ipynb.
- import batch.ipynb notebook is a tool which generates nodes and edges as distinct CSV files, which then can be imported into Neo4j.
- live import.ipynb notebook is a tool which creates tEKGs in real-time using a Cypher query language.
- hybrid import.ipynb nootebook is a combined approach of batch and live mode.
Depending on the mode you wish to use for the transformation, open the corresponding notebook. For instance, if you choose to execute the batch mode tool, open the import batch.ipynb notebook. Before running any cells, in the setup cell of the notebook, specify the experiment_name. Assuming you want to transform the file runningExample-course.jsonocel from the ocel2 directory, you need to set the experiment_name as follows:
experiment_name = 'runningExample-course'Once import batch.ipynb finished, some CSV files are generated in the export directory, which you can load them into Neo4j as instructed in this link.
For executing live import.ipynb and hybrid-import.ipynb, make sure Neo4j is up before running the code. For setting up Neo4j, you can check the documentation. Additinally, it is important to note that, while creating a docker image, you can utilize the usage of memory. For configuration, you can check the instructions provided in the documentation.
As an evaluation for the implemented tools, we compared the running time for transformation of OCEL 2.0 log to tEKG for each of the modes and created a table matrix, which is demonstrated in the evaluation section of the paper. The experiment results can be found in the directory of experiments. As an instance, batch_runningExample-course.json is the result of the experiment for the running example indicating time for transforming and importing OCEL 2.0 log file into tEKG for runningExample-course.jsonocel.