Language Individuation and Marker Words: Shakespeare and His Maxwell's Demon

Abstract

BackgroundWithin the structural and grammatical bounds of a common language, all authors develop their own distinctive writing styles. Whether the relative occurrence of common words can be measured to produce accurate models of authorship is of particular interest. This work introduces a new score that helps to highlight such variations in word occurrence, and is applied to produce models of authorship of a large group of plays from the Shakespearean era.MethodologyA text corpus containing 55,055 unique words was generated from 168 plays from the Shakespearean era (16th and 17th centuries) of undisputed authorship. A new score, CM1, is introduced to measure variation patterns based on the frequency of occurrence of each word for the authors John Fletcher, Ben Jonson, Thomas Middleton and William Shakespeare, compared to the rest of the authors in the study (which provides a reference of relative word usage at that time). A total of 50 WEKA methods were applied for Fletcher, Jonson and Middleton, to identify those which were able to produce models yielding over 90% classification accuracy. This ensemble of WEKA methods was then applied to model Shakespearean authorship across all 168 plays, yielding a Matthews' correlation coefficient (MCC) performance of over 90%. Furthermore, the best model yielded an MCC of 99%.ConclusionsOur results suggest that different authors, while adhering to the structural and grammatical bounds of a common language, develop measurably distinct styles by the tendency to over-utilise or avoid particular common words and phrasings. Considering language and the potential of words as an abstract chaotic system with a high entropy, similarities can be drawn to the Maxwell's Demon thought experiment; authors subconsciously favour or filter certain words, modifying the probability profile in ways that could reflect their individuality and style.

Clustering Consumers Based on Trust, Confidence and Giving Behaviour: Data-Driven Model Building for Charitable Involvement in the Australian Not-For-Profit Sector

“…In order to highlight the individual and most clearly identifiable characteristics of each cluster generated by the MST-kNN agglomerative algorithm, a new score introduced in the area of Computational Linguistics [ 19 ]; and Bioinformatics [ 20 ] was computed. In other words, the use of the CM1 score in this study is to find the most salient features for each cluster.…”

“…This difference is moderated by the range of values observed in members of all the other clusters, Y , which has a greater set of samples, instead of the combined standard deviation of the specific cluster, X , and all the other clusters together; Y . For specific details of this score we refer to a study published by Marsden et al [ 19 ]. …”

An adaptive memetic algorithm for feature selection using proximity graphs

“…We compared the performance of the most successful setup of FSPMA (MA3 algorithm with MineMink and MST) with nine univariate and multivariate feature selection methods. Among the univariate supervised methods chosen are chi‐square feature selection ( Chi ), information gain ratio ( GainRatio ), information gain ( IG ), ReliefF , symmetrical uncertainty ( SU ) from WEKA, and CM1 score . Among the multivariate supervised methods, we selected CFS , consistency subset feature selection ( Consistency ) from WEKA, and the ( α , β )‐ k ‐feature set .…”

“…Among the univariate supervised methods chosen are chi-square feature selection (Chi), information gain ratio (GainRatio), information gain (IG), ReliefF, symmetrical uncertainty (SU) from WEKA, and CM1 score. 39 Among the multivariate supervised methods, we selected CFS, consistency subset feature selection (Consistency) from WEKA, and the ( , )-k-feature set. 40 We implemented the ( , )-k-feature set and CM1 score methods, and for all other methods, we adopted the WEKA implementation using their default configurations.…”

An Information Theoretic Clustering Approach for Unveiling Authorship Affinities in Shakespearean Era Plays and Poems

“…Researchers have debated the merits of culling word lists according to various rules as opposed to using all the words within a given category [9] , [17] , [18] . In a previous research, we demonstrated that such consideration can potentially reflect the authors' individuality and style [19] . In contrast, this research is carried out considering all the words –including stop words.…”