The overall goal of this project is to detect and link similar hemistichs (half verses), verses and epigrams, which will result in a more dynamic system to connect related epigrams in the Database of Byzantine Book Epigrams (DBBE). To achieve this aim, insights and approaches from two active research lines within natural language processing (NLP) will be investigated, viz. automatic linguistic processing of text and machine learning approaches to measure orthographic and semantic similarity between text strings. In addition, a pilot study will be performed that extends the intra-lingual search for similar epigrams in the medieval Greek DBBE to an inter-lingual search for related epigrams in other languages, and more specifically Latin. To this end, the thriving new NLP research line of cross-lingual embeddings will be investigated.