Automation of repetitive tasks can be achieved with Robotic Process Automation (RPA) using scripts that encode fine-grained interactions with software applications on desktops and the web. Automating these processes can be achieved through several applications. It is possible for users to record desktop activity, including metadata, with these tools. The very fine-grained steps in the processes contain details about very small steps that the user takes. Several steps are involved in this process, including clicking on buttons, typing text, selecting the text, and changing the focus. Automating these processes requires connectors connecting them to the appropriate applications. Currently, users choose these connectors manually rather than automatically being linked to processes. In this thesis, we propose a method for recommending the top-k suitable connectors based on event logs for each process. This method indicates that we can use process discovery, create the process models of the train processes with identified connectors, and calculate the conformance checking between the process models and test event logs (unknown connectors). Then we select top-k maximum values of the conformance checking results and observe that we have the suitable connector with 80% accuracy among the top-3 recommended connectors. This solution can be configurable by changing the parameters and the methods of process discovery and conformance checking.

Automation of repetitive tasks can be achieved with Robotic Process Automation (RPA) using scripts that encode fine-grained interactions with software applications on desktops and the web. Automating these processes can be achieved through several applications. It is possible for users to record desktop activity, including metadata, with these tools. The very fine-grained steps in the processes contain details about very small steps that the user takes. Several steps are involved in this process, including clicking on buttons, typing text, selecting the text, and changing the focus. Automating these processes requires connectors connecting them to the appropriate applications. Currently, users choose these connectors manually rather than automatically being linked to processes. In this thesis, we propose a method for recommending the top-k suitable connectors based on event logs for each process. This method indicates that we can use process discovery, create the process models of the train processes with identified connectors, and calculate the conformance checking between the process models and test event logs (unknown connectors). Then we select top-k maximum values of the conformance checking results and observe that we have the suitable connector with 80% accuracy among the top-3 recommended connectors. This solution can be configurable by changing the parameters and the methods of process discovery and conformance checking.

Recommender System Based on Process Mining

NIKRAFTAR KHAMENE, FATEMEH
2022/2023

Abstract

Automation of repetitive tasks can be achieved with Robotic Process Automation (RPA) using scripts that encode fine-grained interactions with software applications on desktops and the web. Automating these processes can be achieved through several applications. It is possible for users to record desktop activity, including metadata, with these tools. The very fine-grained steps in the processes contain details about very small steps that the user takes. Several steps are involved in this process, including clicking on buttons, typing text, selecting the text, and changing the focus. Automating these processes requires connectors connecting them to the appropriate applications. Currently, users choose these connectors manually rather than automatically being linked to processes. In this thesis, we propose a method for recommending the top-k suitable connectors based on event logs for each process. This method indicates that we can use process discovery, create the process models of the train processes with identified connectors, and calculate the conformance checking between the process models and test event logs (unknown connectors). Then we select top-k maximum values of the conformance checking results and observe that we have the suitable connector with 80% accuracy among the top-3 recommended connectors. This solution can be configurable by changing the parameters and the methods of process discovery and conformance checking.
2022
Recommender System Based on Process Mining
Automation of repetitive tasks can be achieved with Robotic Process Automation (RPA) using scripts that encode fine-grained interactions with software applications on desktops and the web. Automating these processes can be achieved through several applications. It is possible for users to record desktop activity, including metadata, with these tools. The very fine-grained steps in the processes contain details about very small steps that the user takes. Several steps are involved in this process, including clicking on buttons, typing text, selecting the text, and changing the focus. Automating these processes requires connectors connecting them to the appropriate applications. Currently, users choose these connectors manually rather than automatically being linked to processes. In this thesis, we propose a method for recommending the top-k suitable connectors based on event logs for each process. This method indicates that we can use process discovery, create the process models of the train processes with identified connectors, and calculate the conformance checking between the process models and test event logs (unknown connectors). Then we select top-k maximum values of the conformance checking results and observe that we have the suitable connector with 80% accuracy among the top-3 recommended connectors. This solution can be configurable by changing the parameters and the methods of process discovery and conformance checking.
Process Mining
Process Automation
Recommender System
RPA
File in questo prodotto:
File Dimensione Formato  
UNIPD_Thesis_FatemehNikraftarKhamene_PDFA.pdf

accesso aperto

Dimensione 1.55 MB
Formato Adobe PDF
1.55 MB Adobe PDF Visualizza/Apri

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/47691