Transfer learning is one of the prevailing approaches towards training language-specific BERT models. However, some languages have uncommon features that may prove to be challenging to more domain-general models but not domain-specific models. Comparing the performance of BERTić, a Bosnian-Croatian-Montenegrin-Serbian model, and Multilingual BERT on a Named-Entity Recognition (NER) task and Masked Language Modelling (MLM) task based around a rare phenomenon of indeclinable female foreign names in Serbian reveals how the different training approaches impacts their performance. Multilingual BERT is shown to perform better than BERTić in the NER task, but BERTić greatly exceeds in the MLM task. Thus, there are applications both for domain-general training and domain-specific training depending on the tasks at hand.

Collostructional analysis is a technique devised to find correlations between particular words and linguistic constructions in order to analyse meaning associations of these constructions. Contrasting collostructional analysis results with output from BERT might provide insights into the way BERT represents the meaning of linguistic constructions. This study tests to what extent English BERT’s meaning representations correspond to known constructions from the linguistics literature by means of two tasks that we propose. Firstly, by predicting the words that can be used in open slots of constructions, the meaning associations of more lexicalized constructions can be observed. Secondly, by finding similar sequences using BERT’s output embeddings and manually reviewing the resulting sentences, we can observe whether instances of less lexicalized constructions are clustered together in semantic space. These two methods show that BERT represents constructional meaning to a certain extent, but does not separate instances of a construction from a near-synonymous construction that has a different form.

We perform a direct intrinsic evaluation of word embeddings trained on the works of a single philosopher. Six models are compared to human judgements elicited using two tasks: a synonym detection task and a coherence task. We apply a method that elicits judgements based on explicit knowledge from experts, as the linguistic intuition of non-expert participants might differ from that of the philosopher. We find that an in-domain SVD model has the best 1-nearest neighbours for target terms, while transfer learning-based Nonce2Vec performs better for low frequency target terms.

We evaluate the use of direct intrinsic word embedding evaluation tasks for specialized language. Our case study is philosophical text: human expert judgements on the relatedness of philosophical terms are elicited using a synonym detection task and a coherence task. Uniquely for our task, experts must rely on explicit knowledge and cannot use their linguistic intuition, which may differ from that of the philosopher. We find that inter-rater agreement rates are similar to those of more conventional semantic annotation tasks, suggesting that these tasks can be used to evaluate word embeddings of text types for which implicit knowledge may not suffice.

Computational linguistic research on language change through distributional semantic (DS) models has inspired researchers from fields such as philosophy and literary studies, who use these methods for the exploration and comparison of comparatively small datasets traditionally analyzed by close reading. Research on methods for small data is still in early stages and it is not clear which methods achieve the best results. We investigate the possibilities and limitations of using distributional semantic models for analyzing philosophical data by means of a realistic use-case. We provide a ground truth for evaluation created by philosophy experts and a blueprint for using DS models in a sound methodological setup. We compare three methods for creating specialized models from small datasets. Though the models do not perform well enough to directly support philosophers yet, we find that models designed for small data yield promising directions for future work.

We address the problem of creating and evaluating quality Neo-Latin word embeddings for the purpose of philosophical research, adapting the Nonce2Vec tool to learn embeddings from Neo-Latin sentences. This distributional semantic modeling tool can learn from tiny data incrementally, using a larger background corpus for initialization. We conduct two evaluation tasks: definitional learning of Latin Wikipedia terms, and learning consistent embeddings from 18th century Neo-Latin sentences pertaining to the concept of mathematical method. Our results show that consistent Neo-Latin word embeddings can be learned from this type of data. While our evaluation results are promising, they do not reveal to what extent the learned models match domain expert knowledge of our Neo-Latin texts. Therefore, we propose an additional evaluation method, grounded in expert-annotated data, that would assess whether learned representations are conceptually sound in relation to the domain of study.

We present a novel, domain expert-controlled, replicable procedure for the construction of concept-modeling ground truths with the aim of evaluating the application of word embeddings. In particular, our method is designed to evaluate the application of word and paragraph embeddings in concept-focused textual domains, where a generic ontology does not provide enough information. We illustrate the procedure, and validate it by describing the construction of an expert ground truth, QuiNE-GT. QuiNE-GT is built to answer research questions concerning the concept of naturalized epistemology in QUINE, a 2-million-token, single-author, 20th-century English philosophy corpus of outstanding quality, cleaned up and enriched for the purpose. To the best of our ken, expert concept-modeling ground truths are extremely rare in current literature, nor has the theoretical methodology behind their construction ever been explicitly conceptualised and properly systematised. Expert-controlled concept-modeling ground truths are however essential to allow proper evaluation of word embeddings techniques, and increase their trustworthiness in specialised domains in which the detection of concepts through their expression in texts is important. We highlight challenges, requirements, and prospects for future work.

In this work, we address the evaluation of distributional semantic models trained on smaller, domain-specific texts, specifically, philosophical text. Specifically, we inspect the behaviour of models using a pre-trained background space in learning. We propose a measure of consistency which can be used as an evaluation metric when no in-domain gold-standard data is available. This measure simply computes the ability of a model to learn similar embeddings from different parts of some homogeneous data. We show that in spite of being a simple evaluation, consistency actually depends on various combinations of factors, including the nature of the data itself, the model used to train the semantic space, and the frequency of the learnt terms, both in the background space and in the in-domain data of interest.

Certain phenomena of interest to linguists mainly occur in low-resource languages, such as contact-induced language change. We show that it is possible to study contact-induced language change computationally in a historical variety of a low-resource language, Early-Modern Frisian, by creating a model using features that were established to be relevant in a closely related language, modern Dutch. This allows us to test two hypotheses on two types of language contact that may have taken place between Frisian and Dutch during this time. Our model shows that Frisian verb cluster word orders are associated with different context features than Dutch verb orders, supporting the ‘learned borrowing’ hypothesis.

This work investigates the application of a measure of surprisal to modeling a grammatical variation phenomenon between near-synonymous constructions. We investigate a particular variation phenomenon, word order variation in Dutch two-verb clusters, where it has been established that word order choice is affected by processing cost. Several multifactorial corpus studies of Dutch verb clusters have used other measures of processing complexity to show that this factor affects word order choice. This previous work allows us to compare the surprisal measure, which is based on constraint satisfaction theories of language modeling, to those previously used measures, which are more directly linked to empirical observations of processing complexity. Our results show that surprisal does not predict the word order choice by itself, but is a significant predictor when used in a measure of uniform information density (UID). This lends support to the view that human language processing is facilitated not so much by predictable sequences of words but more by sequences of words in which information is spread evenly.