1994/01/01 by Thomas Merriam · 85 citations
Computer Science · #Authorship Attribution and Profiling #Stylometry #Tragedy (event) #Artificial neural network #Set (abstract data type) #Literature #Computer science #Canon #Artificial intelligence #Natural language processing #Art #Programming language
paper · doi:10.1093/llc/9.1.1
published in Literary and Linguistic Computing 9(1), 1-6 (Oxford University Press)
openalex publication_date 1994/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2025/11/06
Using principles set out in an earlier paper, a neural network was constructed to discriminate between the works of Shakespeare and his contemporary Christopher Marlowe. Once trained using works from the core canon of the two dramatists, the network successfully classified works to which it had not been previously exposed. In the light of these favourable results, we used the network to classify a number of anonymous works. Strong support emerged for Tucker Brooke's view that The True Tragedy is the Marlovian original of Henry VI, Part 3, the latter being the product of subsequent revisions by Shakespeare.