Innovation is a knowledge-intensive process where much of the codified knowledge is stored in textual formats, such as patents or scientific papers. As businesses increasingly rely on this codified knowledge to remain competitive and continue to innovate, they face the challenge of extracting value from the rapidly growing volume of textual data. With the digital transformation, this knowledge has become not only more voluminous but also more accessible, presenting both opportunities and challenges for innovation management. NLP plays a pivotal role in analysing this body of unstructured textual data (Antons et al., 2020). With the advent of transformer-based models like Google’s BERT, OpenAI’s GPT, or Meta’s Llama, a novel class of NLP tools – referred to as LLMs – has emerged. LLMs are highly advanced AI systems trained on a vast corpus of documents to “learn” the distribution of words in texts and predict probable word sequences. These models have revolutionized how we analyse and extract innovation knowledge, empowering researchers, innovators, designers, and managers to leverage a firm’s knowledge assets across a wide range of tasks (Giordano et al., 2024). In addition to their analytical potential, LLMs are transforming the social dynamics of innovation work. Their integration into team processes raises important questions about leadership, collaboration, and diversity: how different voices contribute to knowledge creation, how authority and expertise are negotiated between humans and LLMs systems, and how such interactions affect equitable participation in innovation management.

NLP techniques and LLMs enhance innovation management by offering two capabilities: analysing existing knowledge and generating new knowledge (Bouschery et al., 2023). Firstly, they can process vast amounts of existing knowledge enabling firms to convert it into new knowledge that can then be applied to their innovation processes (Just, 2024). Secondly, NLP are not only limited to analysing existing data; they are also capable of generating new knowledge. For instance, in new product development, LLMs can assist in generating novel ideas for product features, suggest improvements, or simulate potential innovations by integrating different sources of information.

This conference track aims to explore the vast potential applications, novel methodologies, and emerging challenges of employing LLMs to guide various stages of the innovation process. It will also focus on other NLP techniques, such as topic modelling, named entity recognition, and sentiment analysis, which are different from LLMs in how they extract knowledge. While traditional NLP methods may not be as advanced as LLMs, they can still perform effectively in specific situations and may offer advantages in terms of cost-efficiency and bias reduction. The track encourages submissions of papers that include, but are not limited to:

• Propose new LLM-based methods to support large-scale data analysis and assist humans across various tasks in the innovation management process;
• Apply traditional NLP techniques or/and LLMs to integrate diverse innovation-related knowledge and uncover patterns that generate novel insights;
• Compare “traditional” NLP methods with LLMs in the context of innovation management;
• Examine how LLMs reshape collaboration, leadership, and diversity dynamics within innovation and R&D teams.