633 research outputs found
The CREENDER Tool for Creating Multimodal Datasets of Images and Comments
No full textWhile text-only datasets are widely produced and used for research purposes, limitations set by image-based social media platforms like Instagram make it difficult for researchers to experiment with multimodal data. We therefore developed CREENDER, an annotation tool to create multimodal datasets with images associated with semantic tags and comments, which we make freely available under Apache 2.0 license. The software has been extensively tested with school classes, allowing us to improve the tool and add useful features not planned in the first development phase.Mentre i dataset testuali sono ampiamenti creati e usati per scopi di ricerca, le limitazioni imposte dai social media basati sulle immagini (come Instagram) rendono difficile per i ricercatori sperimentare con dati multimodali. Abbiamo quindi sviluppato CREENDER, un tool di annotazione per la creazione di dataset multimodali in cui immagini vengono associate a etichette semantiche e commenti, e che abbiamo reso disponibile gratuitamente con la licenza Apache 2.0. Il software è stato testato in un laboratorio con alcune classi scolastiche, permettendoci di ottimizzare alcune procedure e di aggiungere feature non previste nella prima release
Adaptive Complex Word Identification through False Friend Detection
No full textAutomated complex word identification (CWI) is a crucial task in several applications, from readability assessment to lexical simplification. So far, several works have modeled CWI with the goal of targeting the needs of non-native speakers. However, studies in language acquisition show that different native languages can create positive or negative interferences w.r.t. reading comprehension, favouring or hindering the understanding of a document in a foreign language. Therefore, we propose to modify CWI to address the specific difficulties connected to different native languages. In particular, we present a pipeline that, based on the user native language, identifies complex terms by automatically detecting cognates and false friends on the fly. The selection presented by the CWI module is adaptive in that it changes depending on the native language of the user. We implement and evaluate our approach for four different native languages (French, English, German and Spanish), in a setting where documents are written in Italian and should be read by language learners with low proficiency. We show that a personalised strategy based on false friend detection identifies complex terms that are different from those usually selected with standard approaches based on word frequency
BullyFrame: Cyberbullying Meets FrameNet
No full textThis paper presents BullyFrame, a dataset of cyberbulling interactions collected from WhatsApp conversations in Italian and annotated with FrameNet semantic frames. We will describe the creation of the dataset discussing the problematic aspects found in the annotation process, such as the lack of coverage of FrameNet for the annotation of texts extracted from social media. Finally, we present a preliminary study that describes the relations between the frames and the cyberbullying-related annotation of the original datase
A Multimodal Dataset of Images and Text to Study Abusive Language
No full textIn this paper, we present a novel dataset composed of images and comments in Italian, created with teenagers in classes using a simulated scenario to raise awareness on cyberbullying phenomena. Potentially offensive comments have been collected for more than 1,000 images and manually assigned to a semantic category. Our analysis shows that the presence of human subjects, as well as the gender of the people present in the pictures trigger different types of comment, and provides novel insight into the connection between images posted on social media and offensive messages. We also compare our corpus with a similar one obtained with WhatsApp, showing that comments to images show different characteristics compared to text-only interactions
Abuse is Contextual, What about NLP? The Role of Context in Abusive Language Annotation and Detection
No full textThe datasets most widely used for abusive language detection contain lists of
messages, usually tweets, that have been manually judged as abusive or not by
one or more annotators, with the annotation performed at message level. In this
paper, we investigate what happens when the hateful content of a message is
judged also based on the context, given that messages are often ambiguous and
need to be interpreted in the context of occurrence. We first re-annotate part
of a widely used dataset for abusive language detection in English in two
conditions, i.e. with and without context. Then, we compare the performance of
three classification algorithms obtained on these two types of dataset, arguing
that a context-aware classification is more challenging but also more similar
to a real application scenario
BERToldo, the Historical BERT for Italian
No full textRecent works in historical language processing have shown that transformer-based models can be successfully created using historical corpora, and that using them for analysing and classifying data from the past can be beneficial compared to standard transformer models. This has led to the creation of BERT-like models for different languages trained with digital repositories from the past. In this work we introduce the Italian version of historical BERT, which we call BERToldo. We evaluate the model on the task of PoS-tagging Dante Alighieri’s works, considering not only the tagger performance but also the model size and the time needed to train it. We also address the problem of duplicated data, which is rather common for languages with a limited availability of historical corpora. We show that deduplication reduces training time without affecting performance. The model and its smaller versions are all made available to the research community
Diciotto anni: e dopo?
No full textIl capitolo presenta una sintesi dei risultati di un'indagine realizzata a Padova nel 2017 con particolare attenzione per i processi di transizione dei minori stranieri non accompagnati dai contesti di accoglienza alla vita adulta. Dati e risultati vengono messi in relazione con i recenti studi europei in materia
Hepatogenous diabetes. Is it time to separate it from type 2 diabetes?
No full textBy definition, hepatogenous diabetes is directly caused by loss of liver function, implying that it develops after cirrhosis onset. Therefore, it should be distinguished from type 2 diabetes developing before cirrhosis onset, in which specific causes of liver disease play a major role, in addition to traditional risk factors. Currently, although hepatogenous diabetes shows distinct pathophysiological and clinical features, it is not considered as an autonomous entity. Recent evidence suggests that the failing liver exerts an independent "toxic" effect on pancreatic islets resulting in β-cell dysfunction. Moreover, patients with hepatogenous diabetes usually present with normal fasting glucose and haemoglobin A1c levels and abnormal response to an oral glucose tolerance test, which is therefore required for diagnosis. This article discusses the need to separate hepatogenous diabetes from type 2 diabetes occurring in subjects with chronic liver disease and to identify individuals suffering from this condition for prognostic and therapeutic purposes
FBK-DH at SemEval-2020 Task 12: Using Multi-channel BERT for Multilingual Offensive Language Detection
No full textIn this paper we present our submission to sub-task A at SemEval 2020 Task 12: Multilingual Offensive Language Identification in Social Media (OffensEval2). For Danish, Turkish, Arabic and Greek, we develop an architecture based on transfer learning and relying on a two-channel BERT model, in which the English BERT and the multilingual one are combined after creating a machine-translated parallel corpus for each language in the task. For English, instead, we adopt a more standard, single-channel approach. We find that, in a multilingual scenario, with some languages having small training data, using parallel BERT models with machine translated data can give systems more stability, especially when dealing with noisy data. The fact that machine translation on social media data may not be perfect does not hurt the overall classification performance
Agreeing to Disagree: Annotating Offensive Language Datasets with Annotators' Disagreement
No full textSince state-of-the-art approaches to offensive language detection rely on
supervised learning, it is crucial to quickly adapt them to the continuously
evolving scenario of social media. While several approaches have been proposed
to tackle the problem from an algorithmic perspective, so to reduce the need
for annotated data, less attention has been paid to the quality of these data.
Following a trend that has emerged recently, we focus on the level of agreement
among annotators while selecting data to create offensive language datasets, a
task involving a high level of subjectivity. Our study comprises the creation
of three novel datasets of English tweets covering different topics and having
five crowd-sourced judgments each. We also present an extensive set of
experiments showing that selecting training and test data according to
different levels of annotators' agreement has a strong effect on classifiers
performance and robustness. Our findings are further validated in cross-domain
experiments and studied using a popular benchmark dataset. We show that such
hard cases, where low agreement is present, are not necessarily due to
poor-quality annotation and we advocate for a higher presence of ambiguous
cases in future datasets, particularly in test sets, to better account for the
different points of view expressed online.Comment: To appear at EMNLP 2021 (long paper
- …
