This paper presents our participation to the “Human Value Detection shared task (Kiesel et al., 2023), as “Andronicus of Rhodes. We describe the approaches behind each entry in the official evaluation, along with the motivation behind each approach. Our best-performing approach has been based on BERT large, with 4 classification heads, implementing two different classification approaches (with different activation and loss functions), and two different partitioning of the training data, to handle class imbalance. Classification is performed through majority voting. The proposed approach outperforms the BERT baseline, ranking in the upper half of the competition.

In this paper, we present the Ellogon Web Annotation Tool. It is a collaborative, web-based annotation tool built upon the Ellogon infrastructure offering an improved user experience and adaptability to various annotation scenarios by making good use of the latest design practices and web development frameworks. Being in development for many years, this paper describes its current architecture, along with the recent modifications that extend the existing functionalities and the new features that were added. The new version of the tool offers document analytics, annotation inspection and comparison features, a modern UI, and formatted text import (e.g. TEI XML documents, rendered with simple markup). We present two use cases that serve as two examples of different annotation scenarios to demonstrate the new functionalities. An appropriate (user-supplied, XML-based) annotation schema is used for each scenario. The first schema contains the relevant components for representing concepts, moral values, and ideas. The second includes all the necessary elements for annotating argumentative units in a document and their binary relations.