With the rise of the digital age, and in particular with the pandemic situation of these days, there is a need for companies to become more and more efficient and valuable. A lot of investments are made in business process technologies that deal with problems regarding costs, time, and risks. In this thesis work, we deal with recommendations of suitable next activities in order to reduce the remaining time of a business process execution. First, we conducted a comparative analysis between two different families of techniques on the task of recommending the most common next activity. Next, we implemented a Random Forest system and a Reinforcement Learning system able to recommend next activities while minimizing the remaining execution time.

With the rise of the digital age, and in particular with the pandemic situation of these days, there is a need for companies to become more and more efficient and valuable. A lot of investments are made in business process technologies that deal with problems regarding costs, time, and risks. In this thesis work, we deal with recommendations of suitable next activities in order to reduce the remaining time of a business process execution. First, we conducted a comparative analysis between two different families of techniques on the task of recommending the most common next activity. Next, we implemented a Random Forest system and a Reinforcement Learning system able to recommend next activities while minimizing the remaining execution time.

What to do as next activity? On the usage of Machine Learning for Process-aware Recommender Systems

CIUCHE, BIANCA ANDREEA
2021/2022

Abstract

With the rise of the digital age, and in particular with the pandemic situation of these days, there is a need for companies to become more and more efficient and valuable. A lot of investments are made in business process technologies that deal with problems regarding costs, time, and risks. In this thesis work, we deal with recommendations of suitable next activities in order to reduce the remaining time of a business process execution. First, we conducted a comparative analysis between two different families of techniques on the task of recommending the most common next activity. Next, we implemented a Random Forest system and a Reinforcement Learning system able to recommend next activities while minimizing the remaining execution time.
2021
What to do as next activity? On the usage of Machine Learning for Process-aware Recommender Systems
With the rise of the digital age, and in particular with the pandemic situation of these days, there is a need for companies to become more and more efficient and valuable. A lot of investments are made in business process technologies that deal with problems regarding costs, time, and risks. In this thesis work, we deal with recommendations of suitable next activities in order to reduce the remaining time of a business process execution. First, we conducted a comparative analysis between two different families of techniques on the task of recommending the most common next activity. Next, we implemented a Random Forest system and a Reinforcement Learning system able to recommend next activities while minimizing the remaining execution time.
Recommender Systems
Process Mining
ReinforcementLearnig
MarkovDecisionProces
File in questo prodotto:
File Dimensione Formato  
Ciuche_Bianca_Andreea.pdf

accesso riservato

Dimensione 2.28 MB
Formato Adobe PDF
2.28 MB Adobe PDF

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/20.500.12608/9963