From observation to insight: Computational abduction and its applications in sustainable strategy research
Kimpimäki, Jaan-Pauli (2023-12-08)
Väitöskirja
08.12.2023
Lappeenranta-Lahti University of Technology LUT
Acta Universitatis Lappeenrantaensis
School of Business and Management
School of Business and Management, Kauppatieteet
Kaikki oikeudet pidätetään.
Julkaisun pysyvä osoite on
https://urn.fi/URN:ISBN:978-952-412-005-0
https://urn.fi/URN:ISBN:978-952-412-005-0
Tiivistelmä
The term computational abduction broadly refers to the use of computational methods in combination with patterns of abductive reasoning. This dissertation focuses on the application of computational abduction in management and organisation studies, addressing three primary research questions: (1) How can computational abductive methods be applied to management and organisation studies? (2) How can we evaluate the quality of outputs from such models? (3) What underlying theoretical bases can help us better understand these outputs?
To explore these questions, this dissertation investigates two sub-classes of computational abduction: similarity-based abduction and mixed-membership modelling. These classes involve the use of both network-based methods, such as pathfinder networks and exponential random graph models, and text-mining approaches such as topic models and co-word analysis.
By linking abductive reasoning with computational models, this research suggests pathways towards more systematic, transparent and rigorous ways of generating hypotheses examining complex organisational phenomena in greater detail and providing specific applications in the analysis of sustainable strategy patterns with applicability in broader management and organisation studies. The key contributions of this work involve embracing integrative modelling, which combines explanation and prediction, and leveraging the power of generative models to create plausible explanations. Furthermore, this dissertation contributes to the understanding of the role of domain theories in applied computational contexts, which help restrict the range of potential explanations.
In summary, this dissertation presents a novel approach to management and organization studies through the application of computational abduction. By combining the strengths of abductive reasoning and computational models, this research opens new avenues for future investigation and proposes little-explored directions toward advancing the field of management and organisation studies.
To explore these questions, this dissertation investigates two sub-classes of computational abduction: similarity-based abduction and mixed-membership modelling. These classes involve the use of both network-based methods, such as pathfinder networks and exponential random graph models, and text-mining approaches such as topic models and co-word analysis.
By linking abductive reasoning with computational models, this research suggests pathways towards more systematic, transparent and rigorous ways of generating hypotheses examining complex organisational phenomena in greater detail and providing specific applications in the analysis of sustainable strategy patterns with applicability in broader management and organisation studies. The key contributions of this work involve embracing integrative modelling, which combines explanation and prediction, and leveraging the power of generative models to create plausible explanations. Furthermore, this dissertation contributes to the understanding of the role of domain theories in applied computational contexts, which help restrict the range of potential explanations.
In summary, this dissertation presents a novel approach to management and organization studies through the application of computational abduction. By combining the strengths of abductive reasoning and computational models, this research opens new avenues for future investigation and proposes little-explored directions toward advancing the field of management and organisation studies.
Kokoelmat
- Väitöskirjat [1093]