FAQ

Our in-house developed algorithm is a sophisticated tool developed to evaluate the potential business impact of scientific publications. This process is powered by advanced Natural Language Processing (NLP) and a suite of machine learning and deep learning models, designed to meticulously analyze and extract meaningful insights from the text data of publications.

1. Data Extraction and Preprocessing: The algorithm begins with the extraction of raw text data from key sections of a publication, such as the abstract, introduction, and conclusion. This step is crucial for understanding the core content and intent of the research, providing a foundation for further analysis.

2. Domain Identification: Identifying the domain of the publication is a critical early step. The algorithm uses NLP techniques to categorize the paper into specific scientific or technological fields. This classification helps tailor subsequent analysis to the context and norms of the domain, enhancing accuracy and relevance.

3. Assessment of Business Potential: The core of the BPS assessment lies in evaluating the business potential of the publication. The algorithm analyzes the text to identify sentences and phrases that indicate high business impact. This includes looking for innovative ideas, applicability to industry problems, and potential for commercialization. The assessment relies on proprietary models that have been trained on extensive datasets to recognize patterns and features associated with high business potential.

4. Comparison with Patent-Like Content: An innovative aspect of the algorithm is its ability to compare sentences within the publication to those found in published patents. This comparison helps in identifying research that not only has high business potential but also possesses qualities that align with successful patents, indicating readiness for commercial exploitation or further development into tangible products or services.

5. Final BPS Determination: The final BPS is determined through a combinatory model that integrates the outputs from the domain identification, business potential assessment, and patent comparison stages. This model has been refined over years of research and development, leveraging multiple datasets created by experts to train the algorithm for high precision and reliability.

The BPS assessment algorithm is a testament to the intersection of AI, NLP, and domain expertise, crafted to bridge the gap between scientific research and its commercial potential. Through continuous refinement and the incorporation of feedback, the algorithm aims to provide accurate, actionable insights for stakeholders looking to identify and invest in promising scientific endeavours.