Search

The analysis of large data volumes for decision-making has evolved from a sideline to a key driver of economic success in the business world of today. As being particularly relevant for technology-oriented organization, patent intelligence – the retrieval, pre-processing and analysis of patent information – has become a relevant means for organization-relevant decisions. This dissertation sheds light using techniques from machine learning for patent intelligence tasks. After summarizing the current literature streams of patent intelligence and machine learning, four publications outline opportunities and challenges that may arise from using supervised or unsupervised machine learning techniques for patent intelligence. For example, supervised machine learning may guide decision making by reducing noise in predictions, for example. Unsupervised machine learning may be useful to explore latent associations between patents when analyzing computationally challenging patent datasets. However, both techniques impose challenges regarding the complexity of the configuration space as well as the transparency and explainability of their underlying algorithms. Implications of this dissertation offer two trade-offs, i.e. in-house versus external procurement and high performance vs. low explanability, and relevant gaps need being addressed by this dissertation open up avenues of further research.

Nowadays, the increasing number of people moving from the countryside to cities leads to fast urbanization, along with its troubles, in several cities in the world. Fast urbanization and the way to deal with it, how urban innovation or smart city is applied to cope with issues of urbanization, technological drivers of urban innovation and movements of technologies in this field, etc. are motivations for this research. Furthermore, the difference in strategies for developing technologies between developed and leapfrogging countries is also another motivation for this research to observe the way urban innovation grows in such two different kinds of countries. By using patent analysis, a T-DNA approach and patent co-classification analysis, this research shows technological drivers in different system levels in urban innovation in two countries - the USA and Vietnam, and provides some insights into technology convergence in urban innovation based on the analysis of different system levels (vertical convergence) and within the super-system regarding city infrastructures (horizontal convergence). Therefore, the research may help urban planners, managers and analysts in companies in urban areas in both cases of two countries.

This thesis takes an evolutionary theoretical perspective to analyse knowledge dynamics linked to the development of Artificial Intelligence (AI) innovations. An introductory chapter presents the thesis’ overarching theoretical framework, research questions, main findings, contributions, policy implications, and limitations. This chapter is followed by three empirical papers that focus on distinct knowledge dynamics related to the creation of AI innovations. The papers provide a comprehensive overview of the technological development of AI, how this development was reflected in countries’ exploration of this technology, and how it affected the innovative activities of firms that introduced AI innovations. The combined findings from these papers, it is argued, allow a better understanding of how specific technologies and global technological development can impact firms’ ability to learn and innovate.

Innovation has become a key factor for economic growth. However, actors cannot rely only on incremental improvements but also need radical innovations now and then in order to stay competitive. These radical novelties may provide a sustainable competitive advantage and thus long-term economic growth. Thus, due to its catalysing role, radical innovations receive attention by policy makers. In spite of ample research on the drivers of innovation processes, the specific mechanisms underlying the emergence and diffusion of radical innovations remain unresolved. This dissertation takes a first step in closing this research gap. Applying an innovation systems approach, the thesis focuses the analysis on the actors and their relations to each other within different boundaries what, put together, forms the central motor for the emergence and diffusion of radical innovations in an innovation system. The present work highlights that several actors (e.g. patenting organisations, technological frontier firms, universities) are important players regarding radical innovation processes within an innovation system. Patenting firms need certain internal knowledge capabilities for an efficient emergence and diffusion of radical innovations. Moreover, public agencies providing R&D funding play a significant role in supporting radical innovation processes. In terms of relations, this dissertation argues that specific regional context conditions and R&D collaboration enhance interactive learning processes aiming at radical novelty within an innovation system. In particular, cognitive relatedness to regional frontier firms is essential to create radical novelty. With regard to the diffusion of radical innovations, it is important for firms to be related to the overall regional knowledge base. R&D collaboration especially increases the generation of radical innovations when building bridges across boundaries such as organisation types, industrial sectors, regions and regional clusters. Finally, the present work provides relevant implications for future research and derives ramifications for policy makers and managers.

Entrepreneurial success within a technology field often derives either from trade secrets or exclusive rights. Consequently, intellectual properties in terms of patents can be seen as a company’s intangible assets that can prohibit competitors from manufacturing, using, selling, offering for sale, or importing patented products or processes. Many approaches have been elaborated for the analysis of patents to gain strategic insights. Examples are provided by data-driven technology analyses to adjust R&D expenses, plan M&A activities, or create technology roadmaps. However, recent approaches lack both sufficient semantic analysis of patent claims and the provision of scientific-based strategic insights. From a legal perspective, patent claims are the most crucial element within patents because exclusive rights derive from patent claims and their features based on patent law. In particular, recent approaches apply the same natural language processes (NLPs) on title, abstract, description, and claims, even though the legal part of patent claims strongly differs in its wording and syntactic structure. Former approaches neglected claim-specific information, such as dependency level in terms of independent, dependent, and multiple dependent claims, as well as the claim categories (e.g. product, manufacturing process claim) defined in the preamble of each claim. This cumulative dissertation aims to overcome these drawbacks by introducing a claim mining framework. Consequently, untapped opportunities for several application fields within technology analysis are explored. Furthermore, this framework provides claim-specific NLPs in terms of both a claim pre-processing that tags claim’s dependency levels and categories and a claim processing that applies claim-specific stopword filters and term weightings to implement similarity measurements. In doing so, application fields within technology analyses, such as information retrieval, technology forecasting, strategic technology planning, patent quality and value identification, and competitor analysis are addressed. Finally, theoretical and managerial implications are drawn to demonstrate the usefulness of this work.

Over the past few decades, scholars have discovered that service innovation transcends the confines of individual organizations and involves interactions and collaborative activities among a network of actors. However, most of the current innovation studies are based on a conceptual view of firm-centric and linear processes of research and development (R&D) that works well in production-oriented economies but may not be as appropriate when it comes to explaining innovation in service-oriented and networked economies, as they become more prevalent in the digital age. As a result, our current understanding of service innovation processes in multi-actor constellations is limited and requires further attention. This thesis contributes a qualitative study, from the automotive engineering service (AES) industry, to the growing body of scholarly work that studies how service innovation emerges from interactions and collaborative activities that take place among a network of actors. In doing so, this thesis employs a qualitative research design that is informed by four different sources of empirical data: preliminary and in-depth interviews, as well as archival data and field observations. The theoretical viewpoint taken in this thesis is informed by the service-dominant (S-D) logic, which entails a broader and systemic view of service innovation. This view emphasizes the centrality of interactions and considers innovation as a process of reconfiguring the institutionalized rules of resource integration in service ecosystems. Based on this perspective, the key finding is an empirically developed process typology that distinguishes five patterns of service innovation in the AES industry. Pre-existing patterns and concepts within the literature are extended and connected through more detailed descriptions of interactions, roles, and activities. Such an in-depth understanding of different service innovation pathways complements and substantiates recent theoretical debates on the connection between service innovation, institutions, and interactions in service ecosystems. Practitioners from the AES industry benefit from this new knowledge because it helps them to understand how their organization needs to interact and operate with other actors from their network to trigger and pursue specific innovation pathways.