AI-Driven Approaches to Product Feature Prioritization
Lehto, Julius (2024)
2024
Teknis-taloudellinen kandidaattiohjelma - Bachelor's Programme in Business and Technology Management
Johtamisen ja talouden tiedekunta - Faculty of Management and Business
Hyväksymispäivämäärä
2024-12-12
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-2024120110655
https://urn.fi/URN:NBN:fi:tuni-2024120110655
Tiivistelmä
Companies developing technology products must decide which product features to include in their product releases. These decisions must consider customer requirements, the feature's value and cost, and the release schedule. Due to limited resources, it is not possible to develop all potential features. Thus, firms must prioritize some features over others. Proper feature prioritization is essential for technologically complex products, where development is typically expensive and time-consuming.
This bachelor’s thesis explores how artificial intelligence can enhance the feature prioritization process and examines the benefits and challenges of AI-based methods. It evaluates five different machine learning techniques. The study is conducted as a systematic literature review, analyzing the applicability of machine learning methods to multicriteria prioritization tasks.
Machine learning enables programs to learn and evolve based on input data. While the use of machine learning in software development has been studied for several years, research specifically addressing product management and product feature prioritization remains limited. This thesis applies findings from software development, particularly in requirements prioritization, to product feature prioritization.
The results show that machine learning offers advantages over manual prioritization methods, especially when working with large datasets. Machine learning methods can reduce human error and the effort required for prioritization. However, these methods also introduce challenges. Developing and using machine learning models requires computational power, and the accuracy of the models depends heavily on the quality of the data used. Additionally, the results produced by these methods are less transparent than those of traditional approaches. Companies must carefully assess the decision's criticality and ability to address the mentioned challenges. When implemented correctly, machine learning can provide an efficient and accurate tool for multicriteria prioritization.
This bachelor’s thesis explores how artificial intelligence can enhance the feature prioritization process and examines the benefits and challenges of AI-based methods. It evaluates five different machine learning techniques. The study is conducted as a systematic literature review, analyzing the applicability of machine learning methods to multicriteria prioritization tasks.
Machine learning enables programs to learn and evolve based on input data. While the use of machine learning in software development has been studied for several years, research specifically addressing product management and product feature prioritization remains limited. This thesis applies findings from software development, particularly in requirements prioritization, to product feature prioritization.
The results show that machine learning offers advantages over manual prioritization methods, especially when working with large datasets. Machine learning methods can reduce human error and the effort required for prioritization. However, these methods also introduce challenges. Developing and using machine learning models requires computational power, and the accuracy of the models depends heavily on the quality of the data used. Additionally, the results produced by these methods are less transparent than those of traditional approaches. Companies must carefully assess the decision's criticality and ability to address the mentioned challenges. When implemented correctly, machine learning can provide an efficient and accurate tool for multicriteria prioritization.
Kokoelmat
- Kandidaatintutkielmat [11031]