Scientific literature encodes a wealth of knowledge relevant to various users. However, the complexity of scientific jargon makes it inaccessible to all but domain specialists. It would be helpful for different types of people to be able to get at least a gist of a paper. Biomedical practitioners often find it difficult to keep up with the information load; but even lay people would benefit from scientific information, for example to dispel medical misconceptions. Besides, in many countries, familiarity with English is limited, let alone scientific English, even among professionals. All this points to the need for simplified access to the scientific literature. We thus present an application aimed at solving this problem, which is capable of summarising scientific text in a way that is tailored to specific types of users, and in their native language. For this objective, we used an LLM that our system queries using user-selected parameters. We conducted an informal evaluation of this prototype using a questionnaire in 3 different languages.

We introduce BUST, a comprehensive benchmark designed to evaluate detectors of texts generated by instruction-tuned large language models (LLMs). Unlike previous benchmarks, our focus lies on evaluating the performance of detector systems, acknowledging the inevitable influence of the underlying tasks and different LLM generators. Our benchmark dataset consists of 25K texts from humans and 7 LLMs responding to instructions across 10 tasks from 3 diverse sources. Using the benchmark, we evaluated 5 detectors and found substantial performance variance across tasks. A meta-analysis of the dataset characteristics was conducted to guide the examination of detector performance. The dataset was analyzed using diverse metrics assessing linguistic features like fluency and coherence, readability scores, and writer attitudes, such as emotions, convincingness, and persuasiveness. Features impacting detector performance were investigated with surrogate models, revealing emotional content in texts enhanced some detectors, yet the most effective detector demonstrated consistent performance, irrespective of writer’s attitudes and text styles. Our approach focused on investigating relationships between the detectors’ performance and two key factors: text characteristics and LLM generators. We believe BUST will provide valuable insights into selecting detectors tailored to specific text styles and tasks and facilitate a more practical and in-depth investigation of detection systems for LLM-generated text.

Hospital discharge letters are a fundamental component of patient management, as they provide the crucial information needed for patient post-hospital care. However their creation is very demanding and resource intensive, as it requires consultation of several reports documenting the patient’s journey throughout their hospital stay. Given the increasing pressures on doctor’s time, tools that can draft a reasonable discharge summary, to be then reviewed and finalized by the experts, would be welcome. In this paper we present a comparative study exploring the possibility of automatic generation of discharge summaries within the context of an hospital in an Italian-speaking region and discuss quantitative and qualitative results. Despite some shortcomings, the obtained results show that a generic generative system such as ChatGPT is capable of producing discharge summaries which are relatively close to the human generated ones, even in Italian.

We report the approaches submitted to the NADI 2024 Subtask 1: Multi-label country-level Dialect Identification (MLDID). The core part was to adapt the information from multi-class data for a multi-label dialect classification task. We experimented with supervised and unsupervised strategies to tackle the task in this challenging setting. Under the supervised setup, we used the model trained using NADI 2023 data and devised approaches to convert the multi-class predictions to multi-label by using information from the confusion matrix or using calibrated probabilities. Under unsupervised settings, we used the Arabic-based sentence encoders and multilingual cross-encoders to retrieve similar samples from the training set, considering each test input as a query. The associated labels are then assigned to the input query. We also tried different variations, such as co-occurring dialects derived from the provided development set. We obtained the best validation performance of 48.5% F-score using one of the variations with an unsupervised approach and the same approach yielded the best test result of 43.27% (Ranked 2).

Recent transformer-based models have made significant strides in generating radiology reports from chest X-ray images. However, a prominent challenge remains; these models often lack prior knowledge, resulting in the generation of synthetic reports that mistakenly reference non-existent prior exams. This discrepancy can be attributed to a knowledge gap between radiologists and the generation models. While radiologists possess patient-specific prior information, the models solely receive X-ray images at a specific time point. To tackle this issue, we propose a novel approach that leverages a rule-based labeler to extract comparison prior information from radiology reports. This extracted comparison prior is then seamlessly integrated into state-of-the-art transformer-based models, enabling them to produce more realistic and comprehensive reports. Our method is evaluated on English report datasets, such as IU X-ray and MIMIC-CXR. The results demonstrate that our approach surpasses baseline models in terms of natural language generation metrics. Notably, our model generates reports that are free from false references to non-existent prior exams, setting it apart from previous models. By addressing this limitation, our approach represents a significant step towards bridging the gap between radiologists and generation models in the domain of medical report generation.

Pre-trained models usually come with a pre-defined tokenization and little flexibility as to what subword tokens can be used in downstream tasks. This problem concerns especially multilingual NLP and low-resource languages, which are typically processed using cross-lingual transfer. In this paper, we aim to find out if the right granularity of tokenization is helpful for a text classification task, namely dialect classification. Aiming at generalizations beyond the studied cases, we look for the optimal granularity in four dialect datasets, two with relatively consistent writing (one Arabic and one Indo-Aryan set) and two with considerably inconsistent writing (one Arabic and one Swiss German set). To gain more control over subword tokenization and ensure direct comparability in the experimental settings, we train a CNN classifier from scratch comparing two subword tokenization methods (Unigram model and BPE). For reference, we compare the results obtained in our analysis to the state of the art achieved by fine-tuning pre-trained models. We show that models trained from scratch with an optimal tokenization level perform better than fine-tuned classifiers in the case of highly inconsistent writing. In the case of relatively consistent writing, fine-tuned models remain better regardless of the tokenization level.

This paper deals with the problem of incre-mental dialect identification. Our goal is toreliably determine the dialect before the fullutterance is given as input. The major partof the previous research on dialect identification has been model-centric, focusing on performance. We address a new question: How much input is needed to identify a dialect? Ourapproach is a data-centric analysis that resultsin general criteria for finding the shortest inputneeded to make a plausible guess. Workingwith three sets of language dialects (Swiss German, Indo-Aryan and Arabic languages), weshow that it is possible to generalize across dialects and datasets with two input shorteningcriteria: model confidence and minimal inputlength (adjusted for the input type). The sourcecode for experimental analysis can be found atGithub.

In this paper, we describe our systems submitted to the NADI Subtask 1: country-wise dialect classifications. We designed two types of solutions. The first type is convolutional neural network CNN) classifiers trained on subword segments of optimized lengths. The second type is fine-tuned classifiers with BERT-based language specific pre-trained models. To deal with the missing dialects in one of the test sets, we experimented with binary classifiers, analyzing the predicted probability distribution patterns and comparing them with the development set patterns. The better performing approach on the development set was fine-tuning language specific pre-trained model (best F-score 26.59%). On the test set, on the other hand, we obtained the best performance with the CNN model trained on subword tokens obtained with a Unigram model (the best F-score 26.12%). Re-training models on samples of training data simulating missing dialects gave the maximum performance on the test set version with a number of dialects lesser than the training set (F-score 16.44%)

This paper describes our submissions to the Social Media Mining for Health Applications (SMM4H) shared task 2022. Our team (mattica) participated in detecting stances and premises in tweets about health mandates related to COVID-19 (Task 2). Our approach was based on using an in-domain Pretrained Language Model, which we fine-tuned by combining different strategies such as leveraging an additional stance detection dataset through two-stage fine-tuning, joint-learning Stance and Premise detection objectives; and ensembling the sentiment-polarity given by an off-the-shelf fine-tuned model.

We describe our submissions to the 6th edition of the Social Media Mining for Health Applications (SMM4H) shared task. Our team (OGNLP) participated in the sub-task: Classification of tweets self-reporting potential cases of COVID-19 (Task 5). For our submissions, we employed systems based on auto-regressive transformer models (XLNet) and back-translation for balancing the dataset.

Negation is a linguistic universal that poses difficulties for cognitive and computational processing. Despite many advances in text analytics, negation resolution remains an acute and continuously researched question in Natural Language Processing. Reliable negation parsing affects results in biomedical text mining, sentiment analysis, machine translation, and many other fields. The availability of multilingual pre-trained general representation models makes it possible to experiment with negation detection in languages that lack annotated data. In this work we test the performance of two state-of-the-art contextual representation models, Multilingual BERT and XLM-RoBERTa. We resolve negation scope by conducting zero-shot transfer between English, Spanish, French, and Russian. Our best result amounts to a token-level F1-score of 86.86% between Spanish and Russian. We correlate these results with a linguistic negation typology and lexical capacity of the models.

Lexical semantic change detection (also known as semantic shift tracing) is a task of identifying words that have changed their meaning over time. Unsupervised semantic shift tracing, focal point of SemEval2020, is particularly challenging. Given the unsupervised setup, in this work, we propose to identify clusters among different occurrences of each target word, considering these as representatives of different word meanings. As such, disagreements in obtained clusters naturally allow to quantify the level of semantic shift per each target word in four target languages. To leverage this idea, clustering is performed on contextualized (BERT-based) embeddings of word occurrences. The obtained results show that our approach performs well both measured separately (per language) and overall, where we surpass all provided SemEval baselines.

The COVID-19 pandemic has been accompanied by such an explosive increase in media coverage and scientific publications that researchers find it difficult to keep up. We are presenting a publicly available pipeline to perform named entity recognition and normalisation in parallel to help find relevant publications and to aid in downstream NLP tasks such as text summarisation. In our approach, we are using a dictionary-based system for its high recall in conjunction with two models based on BioBERT for their accuracy. Their outputs are combined according to different strategies depending on the entity type. In addition, we are using a manually crafted dictionary to increase performance for new concepts related to COVID-19. We have previously evaluated our work on the CRAFT corpus, and make the output of our pipeline available on two visualisation platforms.

Social media platforms offer extensive information about the development of the COVID-19 pandemic and the current state of public health. In recent years, the Natural Language Processing community has developed a variety of methods to extract health-related information from posts on social media platforms. In order for these techniques to be used by a broad public, they must be aggregated and presented in a user-friendly way. We have aggregated ten methods to analyze tweets related to the COVID-19 pandemic, and present interactive visualizations of the results on our online platform, the COVID-19 Twitter Monitor. In the current version of our platform, we offer distinct methods for the inspection of the dataset, at different levels: corpus-wide, single post, and spans within each post. Besides, we allow the combination of different methods to enable a more selective acquisition of knowledge. Through the visual and interactive combination of various methods, interconnections in the different outputs can be revealed.

As our submission to the CRAFT shared task 2019, we present two neural approaches to concept recognition. We propose two different systems for joint named entity recognition (NER) and normalization (NEN), both of which model the task as a sequence labeling problem. Our first system is a BiLSTM network with two separate outputs for NER and NEN trained from scratch, whereas the second system is an instance of BioBERT fine-tuned on the concept-recognition task. We exploit two strategies for extending concept coverage, ontology pretraining and backoff with a dictionary lookup. Our results show that the backoff strategy effectively tackles the problem of unseen concepts, addressing a major limitation of the chosen design. In the cross-system comparison, BioBERT proves to be a strong basis for creating a concept-recognition system, although some entity types are predicted more accurately by the BiLSTM-based system.

We describe our submissions to the 4th edition of the Social Media Mining for Health Applications (SMM4H) shared task. Our team (UZH) participated in two sub-tasks: Automatic classifications of adverse effects mentions in tweets (Task 1) and Generalizable identification of personal health experience mentions (Task 4). For our submissions, we exploited ensembles based on a pre-trained language representation with a neural transformer architecture (BERT) (Tasks 1 and 4) and a CNN-BiLSTM(-CRF) network within a multi-task learning scenario (Task 1). These systems are placed on top of a carefully crafted pipeline of domain-specific preprocessing steps.

Our team at the University of Zürich participated in the first 3 of the 4 sub-tasks at the Social Media Mining for Health Applications (SMM4H) shared task. We experimented with different approaches for text classification, namely traditional feature-based classifiers (Logistic Regression and Support Vector Machines), shallow neural networks, RCNNs, and CNNs. This system description paper provides details regarding the different system architectures and the achieved results.

Author name disambiguation (AND) in publication and citation resources is a well-known problem. Often, information about email address and other details in the affiliation is missing. In cases where such information is not available, identifying the authorship of publications becomes very challenging. Consequently, there have been attempts to resolve such cases by utilizing external resources as references. However, such external resources are heterogeneous and are not always reliable regarding the correctness of information. To solve the AND task, especially when information about an author is not complete we suggest the use of new features such as journal descriptors (JD) and semantic types (ST). The evaluation of different feature models shows that their inclusion has an impact equivalent to that of other important features such as email address. Using such features we show that our system outperforms the state of the art.

We present the first version of a corpus annotated for psychiatric disorders and their etiological factors. The paper describes the choice of text, annotated entities and events/relations as well as the annotation scheme and procedure applied. The corpus is featuring a selection of focus psychiatric disorders including depressive disorder, anxiety disorder, obsessive-compulsive disorder, phobic disorders and panic disorder. Etiological factors for these focus disorders are widespread and include genetic, physiological, sociological and environmental factors among others. Etiological events, including annotated evidence text, represent the interactions between their focus disorders and their etiological factors. Additionally to these core events, symptomatic and treatment events have been annotated. The current version of the corpus includes 175 scientific abstracts. All entities and events/relations have been manually annotated by domain experts and scores of inter-annotator agreement are presented. The aim of the corpus is to provide a first gold standard to support the development of biomedical text mining applications for the specific area of mental disorders which belong to the main contributors to the contemporary burden of disease.

We show how to use large biomedical databases in order to obtain a gold standard for training a machine learning system over a corpus of biomedical text. As an example we use the Comparative Toxicogenomics Database (CTD) and describe by means of a short case study how the obtained data can be applied. We explain how we exploit the structure of the database for compiling training material and a testset. Using a Naive Bayes document classification approach based on words, stem bigrams and MeSH descriptors we achieve a macro-average F-score of 61% on a subset of 8 action terms. This outperforms a baseline system based on a lookup of stemmed keywords by more than 20%. Furthermore, we present directions of future work, taking the described system as a vantage point. Future work will be aiming towards a weakly supervised system capable of discovering complete biomedical interactions and events.

We give an overview of our approach to the extraction of interactions between pharmacogenomic entities like drugs, genes and diseases and suggest classes of interaction types driven by data from PharmGKB and partly following the top level ontology WordNet and biomedical types from BioNLP. Our text mining approach to the extraction of interactions is based on syntactic analysis. We use syntactic analyses to explore domain events and to suggest a set of interaction labels for the pharmacogenomics domain.

We describe techniques for the automatic detection of relationships among domain entities (e.g. genes, proteins, diseases) mentioned in the biomedical literature. Our approach is based on the adaptive selection of candidate interactions sentences, which are then parsed using our own dependency parser. Specific syntax-based filters are used to limit the number of possible candidate interacting pairs. The approach has been implemented as a demonstrator over a corpus of 2000 richly annotated MedLine abstracts, and later tested by participation to a text mining competition. In both cases, the results obtained have proved the adequacy of the proposed approach to the task of interaction detection.

One of the major obstacles for knowledge management remains MultiWord Terminology (MWT). This paper explores the difficulties that arise and describes real world solutions implemented as part of the Parmenides project. Parmenides is being built as an integrated knowledge management package that combines information, MWT and ontology extraction methods in a semi-automated framework. The focus of this paper is on eliciting ontological fragments based on dedicated MWT processing.