Abstract

This paper explores the application of data science to the UK Honours system, focusing on how publicly available information can support due diligence and enhance public confidence. We identify key challenges that must be addressed before such methods can be applied responsibly in this context. We present a data-processing pipeline that collects open-source information via web scraping, extracts relevant text using coreference resolution, and applies sentiment analysis to identify potentially salient evidence about candidates. We evaluate two established sentiment analysis algorithms (afinn, vader) and introduce a novel method, minos, tailored to this use case. Our results indicate that sentence-level filtering and domain-specific sentiment modelling improve the identification of relevant positive and negative signals. The proposed system is intended to augment, rather than replace, human judgement by highlighting potentially high-impact information for further review. This approach may help reduce the risk of oversight in candidate assessment and can be extended to other awards and recognition processes.