Search

We present and make available MedLatinEpi and MedLatinLit , two datasets of medieval Latin texts to be used in research on computational authorship analysis. MedLatinEpi and MedLatinLit consist of 294 and 30 curated texts, respectively, labelled by author; MedLatinEpi texts are of epistolary nature, while MedLatinLit texts consist of literary comments and treatises about various subjects. As such, these two datasets lend themselves to supporting research in authorship analysis tasks, such as authorship attribution, authorship verification, or same-author verification. Along with the datasets, we provide experimental results, obtained on these datasets, for the authorship verification task, i.e., the task of predicting whether a text of unknown authorship was written by a candidate author. We also make available the source code of the authorship verification system we have used, thus allowing our experiments to be reproduced, and to be used as baselines, by other researchers. We also describe the application of the above authorship verification system, using these datasets as training data, for investigating the authorship of two medieval epistles whose authorship has been disputed by scholars.

This chapter looks at other aspects of the “quantification landscape” that have not been covered in the previous chapters, and discusses the evolution of quantification research, from its beginnings to the most recent quantification-based “shared tasks”; the landscape of quantification-based, publicly available software libraries; visualization tools specifically oriented to displaying the results of quantification-based experiments; and other tasks in data science that present important similarities with quantification. This chapter also presents the results of experiments, that we have carried out ourselves, in which we compare many of the methods discussed in Chapter2on a common testing infrastructure.

This chapter sets the stage for the rest of the book by introducing notions fundamental to quantification, such as class proportions, class distributions and their estimation, dataset shift, and the various subtypes of dataset shift which are relevant to the quantification endeavour. In this chapter we also argue why using classification techniques for estimating class distributions is suboptimal, and we then discuss why learning to quantify has evolved as a task of its own, rather than remaining a by-product of classification.

This chapter is possibly the central chapter of the book, and looks at the various supervised learning methods for learning to quantify that have been proposed over the years. These methods belong to two main categories, depending on whether they have an aggregative nature (i.e., they require the classification of all individual unlabelled items as an intermediate step) or a non-aggregative nature (i.e., they perform no classification of individual items). In turn, the aggregative methods may be seen as belonging to two main sub-categories, depending on whether the classification of individual unlabelled items is performed by classifiers trained via general-purpose learners or via special-purpose, quantification-oriented learners.

This chapter provides the motivation for what is to come in the rest of the book by describing the applications that quantification has been put at, ranging from improving classification accuracy in domain adaptation, to measuring and improving the fairness of classification systems with respect to a sensitive attribute, to supporting research and development in fields that are usually more concerned with aggregate data than with individual data, such as the social sciences, political science, epidemiology, market research, ecological modelling, and others.

In this chapter we discuss the experimental evaluation of quantification systems. We look at evaluation measures for the various types of quantification systems (binary, single-label multiclass, multi-label multiclass, ordinal), but also at evaluation protocols for quantification, that essentially consist in ways to extract multiple testing samples for use in quantification evaluation from a single classification test set. The chapter ends with a discussion on how to perform model selection (i.e., hyperparameter optimization) in a quantification-specific way.

In this chapter we look at a number of “advanced” (or niche) topics in quantification, including quantification for ordinal data, “regression quantification” (the task that stands to regression as “standard” quantification stands to classification), cross-lingual quantification for textual data, quantification for networked data, and quantification for streaming data. The chapter ends with a discussion on how to derive confidence intervals for the class prevalence estimates returned by quantification systems.

This chapter concludes the book, discussing possible future developments in the quantification arena.