Search results

Toward live domain-specific languages

Live programming is a style of development characterized by incremental change and immediate feedback. Instead of long edit-compile cycles, developers modify a running program by changing its source code, receiving immediate feedback as it instantly adapts in response. In this paper, we propose an approach to bridge the gap between running programs and textual domain-specific languages (DSLs). The first step of our approach consists of applying a novel model differencing algorithm, tmdiff, to the textual DSL code. By leveraging ordinary text differencing and origin tracking, tmdiff produces deltas defined in terms of the metamodel of a language. In the second step of our approach, the model deltas are applied at run time to update a running system, without having to restart it. Since the model deltas are derived from the static source code of the program, they are unaware of any run-time state maintained during model execution. We therefore propose a generic, dynamic patch architecture, rmpatch, which can be customized to cater for domain-specific state migration. We illustrate rmpatch in a case study of a live programming environment for a simple DSL implemented in Rascal for simultaneously defining and executing state machines.

Collaboration between Speech and Language Therapists and Parents of Children with Developmental Language Disorders

In order to optimize collaboration between Speech and Language Therapists (SLTs) and parents of children with Developmental Language Disorders (DLD), our aim was to study what is needed for SLTs to transition from the parent-as-therapist aide model to the FCC model and optimal collaborate with parents. Chapter 2 discusses the significance of demystifying collaborative working by making explicit how collaboration works. Chapter 3 examines SLTs’ perspectives on engaging parents in parent-child interaction therapy, utilizing a secondary analysis of interview data. Chapter 4 presents a systematic review of specific strategies that therapists can employ to enhance their collaboration with parents of children with developmental disabilities. Chapter 5 explores the needs of parents in their collaborative interactions with SLTs during therapy for their children with DLD, based on semi-structured interviews. Chapter 6 reports the findings from a behavioral analysis of how SLTs currently engage with parents of children with DLD, using data from focus groups. Chapter 7 offers a general discussion on the findings of this thesis, synthesizing insights from previous chapters to propose recommendations for practice and future research.

An Exploration of LLM-Guided Detection of Discursive Patterns in Dutch Social Media

This paper presents a generative large language model (LLM)- guided approach to detect discursive patterns in Dutch social media. Newsrooms of municipalities and public organizations follow public debate on social media to be aware of and prepare for local and global issues and rumours. The onset of these issues and rumours are now detected by communication specialists in newsrooms. Using discourse analysis, we can ground their findings in theory. Devices from discursive psychology such as emotional evaluations are the lowest level components that can help spot and understand issues[ 1]. Thus, a rule-based NLP approach was developed to highlight these devices in a learning environment1. As a next step, we compare the rule-based approach to a large language modeling approach in order to assess the risks and benefits of both methods. We analyze the detection of two discursive patterns in Dutch tweets: magnifying (exaggerated) language use and assigning negative labels to persons or organizations. We compare the generated responses from two Dutch conversational LLMs finetuned on the Dutch language - Geitje-7B-Ultra and Fietje-2-Chat - and a rule-based NLP baseline using a two-fold evaluation process. The results show mixed performance, with the highest performing LLM setups yielding an accuracy of 64% for the maximizing language category and 73% for the negative labels to organizations/persons category. In comparison, the rule-based algorithm achieves an accuracy of 68% for both categories. Although the LLMs perform well in precision, they frequently find patterns in examples where no discursive markers were annotated. Moreover, the rationale analysis shows relatively poor results, pertaining to multiple factors including model size and interpretation of instructions. The results indicate that although there is merit in conducting discursive analysis using generative language models, it comes with the above risks. Recommendations for future work include combining the usage of language models with the rule-based setup for more robust detection as well as further indication of guidelines to improve upon the reasoning process.

Exploring Bias in Data and Models for Misinformation Detection from Text

With the proliferation of misinformation on the web, automatic misinformation detection methods are becoming an increasingly important subject of study. Large language models have produced the best results among content-based methods, which rely on the text of the article rather than the metadata or network features. However, finetuning such a model requires significant training data, which has led to the automatic creation of large-scale misinformation detection datasets. In these datasets, articles are not labelled directly. Rather, each news site is labelled for reliability by an established fact-checking organisation and every article is subsequently assigned the corresponding label based on the reliability score of the news source in question. A recent paper has explored the biases present in one such dataset, NELA-GT-2018, and shown that the models are at least partly learning the stylistic and other features of different news sources rather than the features of unreliable news. We confirm a part of their findings. Apart from studying the characteristics and potential biases of the datasets, we also find it important to examine in what way the model architecture influences the results. We therefore explore which text features or combinations of features are learned by models based on contextual word embeddings as opposed to basic bag-of-words models. To elucidate this, we perform extensive error analysis aided by the SHAP post-hoc explanation technique on a debiased portion of the dataset. We validate the explanation technique on our inherently interpretable baseline model.

Twice random, once mixed: applying mixed models to simultaneously analyze random effects of language and participants

Psychologists, psycholinguists, and other researchers using language stimuli have been struggling for more than 30 years with the problem of how to analyze experimental data that contain two crossed random effects (items and participants). The classical analysis of variance does not apply; alternatives have been proposed but have failed to catch on, and a statistically unsatisfactory procedure of using two approximations (known as F 1 and F 2) has become the standard. A simple and elegant solution using mixed model analysis has been available for 15 years, and recent improvements in statistical software have made mixed models analysis widely available. The aim of this article is to increase the use of mixed models by giving a concise practical introduction and by giving clear directions for undertaking the analysis in the most popular statistical packages. The article also introduces the djmixed add-on package for SPSS, which makes entering the models and reporting their results as straightforward as possible.

Integration of Large Language Models in the Public Sector