The move towards preserving judgement disagreements in NLP requires the identification of adequate evaluation metrics. We identify a set of key properties that such metrics should have, and assess the extent to which natural candidates for soft evaluation such as Cross Entropy satisfy such properties. We employ a theoretical framework, supported by a visual approach, by practical examples, and by the analysis of a real case scenario. Our results indicate that Cross Entropy can result in fairly paradoxical results in some cases, whereas other measures Manhattan distance and Euclidean distance exhibit a more intuitive behavior, at least for the case of binary classification.

We introduce PyRater, an open-source Python toolkit designed for analysing corpora annotations. When creating new annotated language resources, probabilistic models of annotation are the state-of-the-art solution for identifying the best annotators, retrieving the gold standard, and more generally separating annotation signal from noise. PyRater offers a unified interface for several such models and includes an API for the addition of new ones. Additionally, the toolkit has built-in functions to read datasets with multiple annotations and plot the analysis outcomes. In this work, we also demonstrate a novel application of PyRater to zero-shot classifiers, where it effectively selects the best-performing prompt. We make PyRater available to the research community.

Recently, large language models (LLMs) have become increasingly powerful and have become capable of solving a plethora of tasks through proper instructions in natural language. However, the vast majority of testing suites assume that the instructions are written in English, the de facto prompting language. Code intelligence and problem solving still remain a difficult task, even for the most advanced LLMs. Currently, there are no datasets to measure the generalization power for code-generation models in a language other than English. In this work, we present RoCode, a competitive programming dataset, consisting of 2,642 problems written in Romanian, 11k solutions in C, C++ and Python and comprehensive testing suites for each problem. The purpose of RoCode is to provide a benchmark for evaluating the code intelligence of language models trained on Romanian / multilingual text as well as a fine-tuning set for pretrained Romanian models. Through our results and review of related works, we argue for the need to develop code models for languages other than English.

This paper describes the participation of the research laboratory MIND, at the University of Milano-Bicocca, in the SemEval 2023 task related to Learning With Disagreements (Le-Wi-Di). The main goal is to identify the level of agreement/disagreement from a collection of textual datasets with different characteristics in terms of style, language and task. The proposed approach is grounded on the hypothesis that the disagreement between annotators could be grasped by the uncertainty that a model, based on several linguistic characteristics, could have on the prediction of a given gold label.

Large language models have recently risen in popularity due to their ability to perform many natural language tasks without requiring any fine-tuning. In this work, we focus on two novel ideas: (1) generating definitions from examples and using them for zero-shot classification, and (2) investigating how an LLM makes use of the definitions. We thoroughly analyze the performance of GPT-3 model for fine-grained multi-label conspiracy theory classification of tweets using zero-shot labeling. In doing so, we asses how to improve the labeling by providing minimal but meaningful context in the form of the definitions of the labels. We compare descriptive noun phrases, human-crafted definitions, introduce a new method to help the model generate definitions from examples, and propose a method to evaluate GPT-3’s understanding of the definitions. We demonstrate that improving definitions of class labels has a direct consequence on the downstream classification results.

Data Maps (Swayamdipta, et al. 2020) have emerged as a powerful tool for diagnosing large annotated datasets. Given a model fitted on a dataset, these maps show each data instance from the dataset in a 2-dimensional space defined by a) the model’s confidence in the true class and b) the variability of this confidence. In previous work, confidence and variability are usually computed using training dynamics, which requires the fitting of a strong model to the dataset. In this paper, we introduce a novel approach: Zero-Shot Data Maps based on fast bi-encoder networks. For each data point, confidence on the true label and variability are computed over the members of an ensemble of zero-shot models constructed with different — but semantically equivalent — label descriptions, i.e., textual representations of each class in a given label space. We conduct a comparative analysis of maps compiled using traditional training dynamics and our proposed zero-shot models across various datasets. Our findings reveal that Zero-Shot Data Maps generally match those produced by the traditional method while delivering up to a 14x speedup. The code is available [here](https://github.com/symanto-research/zeroshot-cartography).

Cross-lingual transfer learning from high-resource to medium and low-resource languages has shown encouraging results. However, the scarcity of resources in target languages remains a challenge. In this work, we resort to data augmentation and continual pre-training for domain adaptation to improve cross-lingual abusive language detection. For data augmentation, we analyze two existing techniques based on vicinal risk minimization and propose MIXAG, a novel data augmentation method which interpolates pairs of instances based on the angle of their representations. Our experiments involve seven languages typologically distinct from English and three different domains. The results reveal that the data augmentation strategies can enhance few-shot cross-lingual abusive language detection. Specifically, we observe that consistently in all target languages, MIXAG improves significantly in multidomain and multilingual environments. Finally, we show through an error analysis how the domain adaptation can favour the class of abusive texts (reducing false negatives), but at the same time, declines the precision of the abusive language detection model.

High-quality labeled data is paramount to the performance of modern machine learning models. However, annotating data is a time-consuming and costly process that requires human experts to examine large collections of raw data. For conversational agents in production settings with access to large amounts of user-agent conversations, the challenge is to decide what data should be annotated first. We consider the Natural Language Understanding (NLU) component of a conversational agent deployed in a real-world setup with limited resources. We present an active learning pipeline for offline detection of classification errors that leverages two strong classifiers. Then, we perform topic modeling on the potentially mis-classified samples to ease data analysis and to reveal error patterns. In our experiments, we show on a real-world dataset that by using our method to prioritize data annotation we reach 100% of the performance annotating only 36% of the data. Finally, we present an analysis of some of the error patterns revealed and argue that our pipeline is a valuable tool to detect critical errors and reduce the workload of annotators.

Detecting offensive language in under-resourced languages presents a significant real-world challenge for social media platforms. This paper is the first work focused on the issue of offensive language detection in Arabizi, an under-explored topic in an under-resourced form of Arabic. For the first time, a comprehensive and critical overview of the existing work on the topic is presented. In addition, we carry out experiments using different BERT-like models and show the feasibility of detecting offensive language in Arabizi with high accuracy. Throughout a thorough analysis of results, we emphasize the complexities introduced by dialect variations and out-of-domain generalization. We use in our experiments a dataset that we have constructed by leveraging existing, albeit limited, resources. To facilitate further research, we make this dataset publicly accessible to the research community.

The rapid spread of information over social media influences quantitative trading and investments. The growing popularity of speculative trading of highly volatile assets such as cryptocurrencies and meme stocks presents a fresh challenge in the financial realm. Investigating such “bubbles” - periods of sudden anomalous behavior of markets are critical in better understanding investor behavior and market dynamics. However, high volatility coupled with massive volumes of chaotic social media texts, especially for underexplored assets like cryptocoins pose a challenge to existing methods. Taking the first step towards NLP for cryptocoins, we present and publicly release CryptoBubbles, a novel multi- span identification task for bubble detection, and a dataset of more than 400 cryptocoins from 9 exchanges over five years spanning over two million tweets. Further, we develop a set of sequence-to-sequence hyperbolic models suited to this multi-span identification task based on the power-law dynamics of cryptocurrencies and user behavior on social media. We further test the effectiveness of our models under zero-shot settings on a test set of Reddit posts pertaining to 29 “meme stocks”, which see an increase in trade volume due to social media hype. Through quantitative, qualitative, and zero-shot analyses on Reddit and Twitter spanning cryptocoins and meme-stocks, we show the practical applicability of CryptoBubbles and hyperbolic models.

This paper describes our participation in the shared task Fine-Grained Hate Speech Detection on Arabic Twitter at the 5th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT). The shared task is divided into three detection subtasks: (i) Detect whether a tweet is offensive or not; (ii) Detect whether a tweet contains hate speech or not; and (iii) Detect the fine-grained type of hate speech (race, religion, ideology, disability, social class, and gender). It is an effort toward the goal of mitigating the spread of offensive language and hate speech in Arabic-written content on social media platforms. To solve the three subtasks, we employed six different transformer versions: AraBert, AraElectra, Albert-Arabic, AraGPT2, mBert, and XLM-Roberta. We experimented with models based on encoder and decoder blocks and models exclusively trained on Arabic and also on several languages. Likewise, we applied two ensemble methods: Majority vote and Highest sum. Our approach outperformed the official baseline in all the subtasks, not only considering F1-macro results but also accuracy, recall, and precision. The results suggest that the Highest sum is an excellent approach to encompassing transformer output to create an ensemble since this method offered at least top-two F1-macro values across all the experiments performed on development and test data.

Hate speech detection is a prominent and challenging task, since hate messages are often expressed in subtle ways and with characteristics that may vary depending on the author. Hence, many models suffer from the generalization problem. However, retrieving and monitoring hateful content on social media is a current necessity. In this paper, we propose an unsupervised approach using Graph Auto-Encoders (GAE), which allows us to avoid using labeled data when training the representation of the texts. Specifically, we represent texts as nodes of a graph, and use a transformer layer together with a convolutional layer to encode these nodes in a low-dimensional space. As a result, we obtain embeddings that can be decoded into a reconstruction of the original network. Our main idea is to learn a model with a set of texts without supervision, in order to generate embeddings for the nodes: nodes with the same label should be close in the embedding space, which, in turn, should allow us to distinguish among classes. We employ this strategy to detect hate speech in multi-domain and multilingual sets of texts, where our method shows competitive results on small datasets.

Mental disorders are a serious and increasingly relevant public health issue. NLP methods have the potential to assist with automatic mental health disorder detection, but building annotated datasets for this task can be challenging; moreover, annotated data is very scarce for disorders other than depression. Understanding the commonalities between certain disorders is also important for clinicians who face the problem of shifting standards of diagnosis. We propose that transfer learning with linguistic features can be useful for approaching both the technical problem of improving mental disorder detection in the context of data scarcity, and the clinical problem of understanding the overlapping symptoms between certain disorders. In this paper, we target four disorders: depression, PTSD, anorexia and self-harm. We explore multi-aspect transfer learning for detecting mental disorders from social media texts, using deep learning models with multi-aspect representations of language (including multiple types of interpretable linguistic features). We explore different transfer learning strategies for cross-disorder and cross-platform transfer, and show that transfer learning can be effective for improving prediction performance for disorders where little annotated data is available. We offer insights into which linguistic features are the most useful vehicles for transferring knowledge, through ablation experiments, as well as error analysis.

Proactively identifying misinformation spreaders is an important step towards mitigating the impact of fake news on our society. In this paper, we introduce a new contemporary Reddit dataset for fake news spreader analysis, called FACTOID, monitoring political discussions on Reddit since the beginning of 2020. The dataset contains over 4K users with 3.4M Reddit posts, and includes, beyond the users’ binary labels, also their fine-grained credibility level (very low to very high) and their political bias strength (extreme right to extreme left). As far as we are aware, this is the first fake news spreader dataset that simultaneously captures both the long-term context of users’ historical posts and the interactions between them. To create the first benchmark on our data, we provide methods for identifying misinformation spreaders by utilizing the social connections between the users along with their psycho-linguistic features. We show that the users’ social interactions can, on their own, indicate misinformation spreading, while the psycho-linguistic features are mostly informative in non-neural classification settings. In a qualitative analysis we observe that detecting affective mental processes correlates negatively with right-biased users, and that the openness to experience factor is lower for those who spread fake news.

In this paper we present a set of multilingual experiments tackling the task of Stance Detection in five different languages: English, Spanish, Catalan, French and Italian. Furthermore, we study the phenomenon of stance with respect to six different targets – one per language, and two different for Italian – employing a variety of machine learning algorithms that primarily exploit morphological and syntactic knowledge as features, represented throughout the format of Universal Dependencies. Results seem to suggest that the methodology employed is not beneficial per se, but might be useful to exploit the same features with a different methodology.

The paper describes the SemEval-2022 Task 5: Multimedia Automatic Misogyny Identification (MAMI),which explores the detection of misogynous memes on the web by taking advantage of available texts and images. The task has been organised in two related sub-tasks: the first one is focused on recognising whether a meme is misogynous or not (Sub-task A), while the second one is devoted to recognising types of misogyny (Sub-task B). MAMI has been one of the most popular tasks at SemEval-2022 with more than 400 participants, 65 teams involved in Sub-task A and 41 in Sub-task B from 13 countries. The MAMI challenge received 4214 submitted runs (of which 166 uploaded on the leader-board), denoting an enthusiastic participation for the proposed problem. The collection and annotation is described for the task dataset. The paper provides an overview of the systems proposed for the challenge, reports the results achieved in both sub-tasks and outlines a description of the main errors for a comprehension of the systems capabilities and for detailing future research perspectives.

In this paper we describe the systems used by the RoMa team in the shared task on Detecting and Rating Humor and Offense (HaHackathon) at SemEval 2021. Our systems rely on data representations learned through fine-tuned neural language models. Particularly, we explore two distinct architectures. The first one is based on a Siamese Neural Network (SNN) combined with a graph-based clustering method. The SNN model is used for learning a latent space where instances of humor and non-humor can be distinguished. The clustering method is applied to build prototypes of both classes which are used for training and classifying new messages. The second one combines neural language model representations with a linear regression model which makes the final ratings. Our systems achieved the best results for humor classification using model one, whereas for offensive and humor rating the second model obtained better performance. In the case of the controversial humor prediction, the most significant improvement was achieved by a fine-tuning of the neural language model. In general, the results achieved are encouraging and give us a starting point for further improvements.

Hyperpartisan news show an extreme manipulation of reality based on an underlying and extreme ideological orientation. Because of its harmful effects at reinforcing one’s bias and the posterior behavior of people, hyperpartisan news detection has become an important task for computational linguists. In this paper, we evaluate two different approaches to detect hyperpartisan news. First, a text masking technique that allows us to compare style vs. topic-related features in a different perspective from previous work. Second, the transformer-based models BERT, XLM-RoBERTa, and M-BERT, known for their ability to capture semantic and syntactic patterns in the same representation. Our results corroborate previous research on this task in that topic-related features yield better results than style-based ones, although they also highlight the relevance of using higher-length n-grams. Furthermore, they show that transformer-based models are more effective than traditional methods, but this at the cost of greater computational complexity and lack of transparency. Based on our experiments, we conclude that the beginning of the news show relevant information for the transformers at distinguishing effectively between left-wing, mainstream, and right-wing orientations.

Fake news articles often stir the readers’ attention by means of emotional appeals that arouse their feelings. Unlike in short news texts, authors of longer articles can exploit such affective factors to manipulate readers by adding exaggerations or fabricating events, in order to affect the readers’ emotions. To capture this, we propose in this paper to model the flow of affective information in fake news articles using a neural architecture. The proposed model, FakeFlow, learns this flow by combining topic and affective information extracted from text. We evaluate the model’s performance with several experiments on four real-world datasets. The results show that FakeFlow achieves superior results when compared against state-of-the-art methods, thus confirming the importance of capturing the flow of the affective information in news articles.

Eating disorders are a growing problem especially among young people, yet they have been under-studied in computational research compared to other mental health disorders such as depression. Computational methods have a great potential to aid with the automatic detection of mental health problems, but state-of-the-art machine learning methods based on neural networks are notoriously difficult to interpret, which is a crucial problem for applications in the mental health domain. We propose leveraging the power of deep learning models for automatically detecting signs of anorexia based on social media data, while at the same time focusing on interpreting their behavior. We train a hierarchical attention network to detect people with anorexia and use its internal encodings to discover different clusters of anorexia symptoms. We interpret the identified patterns from multiple perspectives, including emotion expression, psycho-linguistic features and personality traits, and we offer novel hypotheses to interpret our findings from a psycho-social perspective. Some interesting findings are patterns of word usage in some users with anorexia which show that they feel less as being part of a group compared to control cases, as well as that they have abandoned explanatory activity as a result of a greater feeling of helplessness and fear.

This paper presents an in-depth investigation of the effectiveness of dependency-based syntactic features on the irony detection task in a multilingual perspective (English, Spanish, French and Italian). It focuses on the contribution from syntactic knowledge, exploiting linguistic resources where syntax is annotated according to the Universal Dependencies scheme. Three distinct experimental settings are provided. In the first, a variety of syntactic dependency-based features combined with classical machine learning classifiers are explored. In the second scenario, two well-known types of word embeddings are trained on parsed data and tested against gold standard datasets. In the third setting, dependency-based syntactic features are combined into the Multilingual BERT architecture. The results suggest that fine-grained dependency-based syntactic information is informative for the detection of irony.

Author profiling studies how language is shared by people. Stylometry techniques help in identifying aspects such as gender, age, native language, or even personality. Author profiling is a problem of growing importance, not only in marketing and forensics, but also in cybersecurity. The aim is not only to identify users whose messages are potential threats from a terrorism viewpoint but also those whose messages are a threat from a social exclusion perspective because containing hate speech, cyberbullying etc. Bots often play a key role in spreading hate speech, as well as fake news, with the purpose of polarizing the public opinion with respect to controversial issues like Brexit or the Catalan referendum. For instance, the authors of a recent study about the 1 Oct 2017 Catalan referendum, showed that in a dataset with 3.6 million tweets, about 23.6% of tweets were produced by bots. The target of these bots were pro-independence influencers that were sent negative, emotional and aggressive hateful tweets with hashtags such as #sonunesbesties (i.e. #theyareanimals). Since 2013 at the PAN Lab at CLEF (https://pan.webis.de/) we have addressed several aspects of author profiling in social media. In 2019 we investigated the feasibility of distinguishing whether the author of a Twitter feed is a bot, while this year we are addressing the problem of profiling those authors that are more likely to spread fake news in Twitter because they did in the past. We aim at identifying possible fake news spreaders as a first step towards preventing fake news from being propagated among online users (fake news aim to polarize the public opinion and may contain hate speech). In 2021 we specifically aim at addressing the challenging problem of profiling haters in social media in order to monitor abusive language and prevent cases of social exclusion in order to combat, for instance, racism, xenophobia and misogyny. Although we already started addressing the problem of detecting hate speech when targets are immigrants or women at the HatEval shared task in SemEval-2019, and when targets are women also in the Automatic Misogyny Identification tasks at IberEval-2018, Evalita-2018 and Evalita-2020, it was not done from an author profiling perspective. At the end of the keynote, I will present some insights in order to stress the importance of monitoring abusive language in social media, for instance, in foreseeing sexual crimes. In fact, previous studies confirmed that a correlation might lay between the yearly per capita rate of rape and the misogynistic language used in Twitter.

This paper describes the system submitted by the PRHLT-UPV team for the task 8 of SemEval-2020: Memotion Analysis. We propose a multimodal model that combines pretrained models of the BERT and VGG architectures. The BERT model is used to process the textual information and VGG the images. The multimodal model is used to classify memes according to the presence of offensive, sarcastic, humorous and motivating content. Also, a sentiment analysis of memes is carried out with the proposed model. In the experiments, the model is compared with other approaches to analyze the relevance of the multimodal model. The results show encouraging performances on the final leaderboard of the competition, reaching good positions in the ranking of systems.

This paper describes the participation of LIMSI_UPV team in SemEval-2020 Task 9: Sentiment Analysis for Code-Mixed Social Media Text. The proposed approach competed in SentiMix HindiEnglish subtask, that addresses the problem of predicting the sentiment of a given Hindi-English code-mixed tweet. We propose Recurrent Convolutional Neural Network that combines both the recurrent neural network and the convolutional network to better capture the semantics of the text, for code-mixed sentiment analysis. The proposed system obtained 0.69 (best run) in terms of F1 score on the given test data and achieved the 9th place (Codalab username: somban) in the SentiMix Hindi-English subtask.

The present paper describes the system submitted by the PRHLT-UPV team for the task 12 of SemEval-2020: OffensEval 2020. The official title of the task is Multilingual Offensive Language Identification in Social Media, and aims to identify offensive language in texts. The languages included in the task are English, Arabic, Danish, Greek and Turkish. We propose a model based on the BERT architecture for the analysis of texts in English. The approach leverages knowledge within a pre-trained model and performs fine-tuning for the particular task. In the analysis of the other languages the Multilingual BERT is used, which has been pre-trained for a large number of languages. In the experiments, the proposed method for English texts is compared with other approaches to analyze the relevance of the architecture used. Furthermore, simple models for the other languages are evaluated to compare them with the proposed one. The experimental results show that the model based on BERT outperforms other approaches. The main contribution of this work lies in this study, despite not obtaining the first positions in most cases of the competition ranking.

The recognition of irony is a challenging task in the domain of Sentiment Analysis, and the availability of annotated corpora may be crucial for its automatic processing. In this paper we describe a fine-grained annotation scheme centered on irony, in which we highlight the tokens that are responsible for its activation, (irony activators) and their morpho-syntactic features. As our case study we therefore introduce a recently released Universal Dependencies treebank for Italian which includes ironic tweets: TWITTIRÒ-UD. For the purposes of this study, we enriched the existing annotation in the treebank, with a further level that includes irony activators. A description and discussion of the annotation scheme is provided with a definition of irony activators and the guidelines for their annotation. This qualitative study on the different layers of annotation applied on the same dataset can shed some light on the process of human annotation, and irony annotation in particular, and on the usefulness of this representation for developing computational models of irony to be used for training purposes.

The paper describes the organization of the SemEval 2019 Task 5 about the detection of hate speech against immigrants and women in Spanish and English messages extracted from Twitter. The task is organized in two related classification subtasks: a main binary subtask for detecting the presence of hate speech, and a finer-grained one devoted to identifying further features in hateful contents such as the aggressive attitude and the target harassed, to distinguish if the incitement is against an individual rather than a group. HatEval has been one of the most popular tasks in SemEval-2019 with a total of 108 submitted runs for Subtask A and 70 runs for Subtask B, from a total of 74 different teams. Data provided for the task are described by showing how they have been collected and annotated. Moreover, the paper provides an analysis and discussion about the participant systems and the results they achieved in both subtasks.

This paper describes the system we developed for SemEval 2019 on Identifying and Categorizing Offensive Language in Social Media (OffensEval - Task 6). The task focuses on offensive language in tweets. It is organized into three sub-tasks for offensive language identification; automatic categorization of offense types and offense target identification. The approach for the first subtask is a deep learning-based ensemble method which uses a Bidirectional LSTM Recurrent Neural Network and a Convolutional Neural Network. Additionally we use the information from part-of-speech tagging of tweets for target identification and combine previous results for categorization of offense types.

This article presents our approach for detecting a target of offensive messages in Twitter, including Individual, Group and Others classes. The model we have created is an ensemble of simpler models, including Logistic Regression, Naive Bayes, Support Vector Machine and the interpolation between Logistic Regression and Naive Bayes with 0.25 coefficient of interpolation. The model allows us to achieve 0.547 macro F1-score.

In the present paper we describe the UPV-28-UNITO system’s submission to the RumorEval 2019 shared task. The approach we applied for addressing both the subtasks of the contest exploits both classical machine learning algorithms and word embeddings, and it is based on diverse groups of features: stylistic, lexical, emotional, sentiment, meta-structural and Twitter-based. A novel set of features that take advantage of the syntactic information in texts is moreover introduced in the paper.

In this paper, we approach the task of native language identification in a realistic cross-corpus scenario where a model is trained with available data and has to predict the native language from data of a different corpus. The motivation behind this study is to investigate native language identification in the Australian academic scenario where a majority of students come from China, Indonesia, and Arabic-speaking nations. We have proposed a statistical embedding representation reporting a significant improvement over common single-layer approaches of the state of the art, identifying Chinese, Arabic, and Indonesian in a cross-corpus scenario. The proposed approach was shown to be competitive even when the data is scarce and imbalanced.

With the uncontrolled increasing of fake news and rumors over the Web, different approaches have been proposed to address the problem. In this paper, we present an approach that combines lexical, word embeddings and n-gram features to detect the stance in fake news. Our approach has been tested on the Fake News Challenge (FNC-1) dataset. Given a news title-article pair, the FNC-1 task aims at determining the relevance of the article and the title. Our proposed approach has achieved an accurate result (59.6 % Macro F1) that is close to the state-of-the-art result with 0.013 difference using a simple feature representation. Furthermore, we have investigated the importance of different lexicons in the detection of the classification labels.

In this paper we describe our participation in the SemEval-2018 task 3 Shared Task on Irony Detection. We have approached the task with our low dimensionality representation method (LDR), which exploits low dimensional features extracted from text on the basis of the occurrence probability of the words depending on each class. Our intuition is that words in ironic texts have different probability of occurrence than in non-ironic ones. Our approach obtained acceptable results in both subtasks A and B. We have performed an error analysis that shows the difference on correct and incorrect classified tweets.

This paper describes an ensemble approach to the SemEval-2018 Task 3. The proposed method is composed of two renowned methods in text classification together with a novel approach for capturing ironic content by exploiting a tailored lexicon for irony detection. We experimented with different ensemble settings. The obtained results show that our method has a good performance for detecting the presence of ironic content in Twitter.

In this paper we describe the system used by the ValenTO team in the shared task on Irony Detection in English Tweets at SemEval 2018. The system takes as starting point emotIDM, an irony detection model that explores the use of affective features based on a wide range of lexical resources available for English, reflecting different facets of affect. We experimented with different settings, by exploiting different classifiers and features, and participated both to the binary irony detection task and to the task devoted to distinguish among different types of irony. We report on the results obtained by our system both in a constrained setting and unconstrained setting, where we explored the impact of using additional data in the training phase, such as corpora annotated for the presence of irony or sarcasm from the state of the art. Overall, the performance of our system seems to validate the important role that affective information has for identifying ironic content in Twitter.

Gender identification in social networks is one of the most popular aspects of user profile learning. Traditionally it has been linked to author profiling, a difficult problem to solve because of the little difference in the use of language between genders. This situation has led to the need of taking into account other information apart from textual data, favoring the emergence of multimodal data. The aim of this paper is to apply neural networks to perform data fusion, using an existing multimodal corpus, the NUS-MSS data set, that (not only) contains text data, but also image and location information. We improved previous results in terms of macro accuracy (87.8%) obtaining the state-of-the-art performance of 91.3%.

The polarity classification task aims at automatically identifying whether a subjective text is positive or negative. When the target domain is different from those where a model was trained, we refer to a cross-domain setting. That setting usually implies the use of a domain adaptation method. In this work, we study the single and cross-domain polarity classification tasks from the string kernels perspective. Contrary to classical domain adaptation methods, which employ texts from both domains to detect pivot features, we do not use the target domain for training. Our approach detects the lexical peculiarities that characterise the text polarity and maps them into a domain independent space by means of kernel discriminant analysis. Experimental results show state-of-the-art performance in single and cross-domain polarity classification.

We present a model to perform authorship attribution of tweets using Convolutional Neural Networks (CNNs) over character n-grams. We also present a strategy that improves model interpretability by estimating the importance of input text fragments in the predicted classification. The experimental evaluation shows that text CNNs perform competitively and are able to outperform previous methods.

Author profiling is the study of how language is shared by people, a problem of growing importance in applications dealing with security, in order to understand who could be behind an anonymous threat message, and marketing, where companies may be interested in knowing the demographics of people that in online reviews liked or disliked their products. In this talk we will give an overview of the PAN shared tasks that since 2013 have been organised at CLEF and FIRE evaluation forums, mainly on age and gender identification in social media, although also personality recognition in Twitter as well as in code sources was also addressed. In 2017 the PAN author profiling shared task addresses jointly gender and language variety identification in Twitter where tweets have been annotated with authors’ gender and their specific variation of their native language: English (Australia, Canada, Great Britain, Ireland, New Zealand, United States), Spanish (Argentina, Chile, Colombia, Mexico, Peru, Spain, Venezuela), Portuguese (Brazil, Portugal), and Arabic (Egypt, Gulf, Levantine, Maghrebi).

We provide several methods for sentence-alignment of texts with different complexity levels. Using the best of them, we sentence-align the Newsela corpora, thus providing large training materials for automatic text simplification (ATS) systems. We show that using this dataset, even the standard phrase-based statistical machine translation models for ATS can outperform the state-of-the-art ATS systems.

Like most of the languages which have only recently started being investigated for the Natural Language Processing (NLP) tasks, Amazigh lacks annotated corpora and tools and still suffers from the scarcity of linguistic tools and resources. The main aim of this paper is to present a new part-of-speech (POS) tagger based on a new Amazigh tag set (AMTS) composed of 28 tags. In line with our goal we have trained Conditional Random Fields (CRFs) to build a POS tagger for the Amazigh language. We have used the 10-fold technique to evaluate and validate our approach. The CRFs 10 folds average level is 87.95% and the best fold level result is 91.18%. In order to improve this result, we have gathered a set of about 8k words with their POS tags. The collected lexicon was used with CRFs confidence measure in order to have a more accurate POS-tagger. Hence, we have obtained a better performance of 93.82%.

The Arabic WordNet project has provided the Arabic Natural Language Processing (NLP) community with the first WordNet-compliant resource. It allowed new possibilities in terms of building sophisticated NLP applications related to this Semitic language. In this paper, we present the new content added to this resource, using semi-automatic techniques, and validated by Arabic native-speaker lexicographers. We also present how this content helps in the implementation of new Arabic NLP applications, especially for Question Answering (QA) systems. The obtained results show the contribution of the added content. The resource, fully transformed into the standard Lexical Markup Framework (LMF), is made available for the community.

Bilingual dictionaries are the key component of the cross-lingual similarity estimation methods. Usually such dictionary generation is accomplished by manual or automatic means. Automatic generation approaches include to exploit parallel or comparable data to derive dictionary entries. Such approaches require large amount of bilingual data in order to produce good quality dictionary. Many time the language pair does not have large bilingual comparable corpora and in such cases the best automatic dictionary is upper bounded by the quality and coverage of such corpora. In this work we propose a method which exploits continuous quasi-comparable corpora to derive term level associations for enrichment of such limited dictionary. Though we propose our experiments for English and Hindi, our approach can be easily extendable to other languages. We evaluated dictionary by manually computing the precision. In experiments we show our approach is able to derive interesting term level associations across languages.

With the constant increase in the amount of information available in online communities, the task of building an appropriate Recommender System to support the user in her decision making process is becoming more and more challenging. In addition to the classical collaborative filtering and content based approaches, taking into account ratings, preferences and demographic characteristics of the users, a new type of Recommender System, based on personality parameters, has been emerging recently. In this paper we describe the TWIN (Tell Me What I Need) Personality Based Recommender System, and report on our experiments and experiences of utilizing techniques which allow the extraction of the personality type from text (following the Big Five model popular in the psychological research). We estimate the possibility of constructing the personality-based Recommender System that does not require users to fill in personality questionnaires. We are applying the proposed system in the online travelling domain to perform TripAdvisor hotels recommendation by analysing the text of user generated reviews, which are freely accessible from the community website.

The simple access to texts on digital libraries and the World Wide Web has led to an increased number of plagiarism cases in recent years, which renders manual plagiarism detection infeasible at large. Various methods for automatic plagiarism detection have been developed whose objective is to assist human experts in the analysis of documents for plagiarism. The methods can be divided into two main approaches: intrinsic and external. Unlike other tasks in natural language processing and information retrieval, it is not possible to publish a collection of real plagiarism cases for evaluation purposes since they cannot be properly anonymized. Therefore, current evaluations found in the literature are incomparable and, very often not even reproducible. Our contribution in this respect is a newly developed large-scale corpus of artificial plagiarism useful for the evaluation of intrinsic as well as external plagiarism detection. Additionally, new detection performance measures tailored to the evaluation of plagiarism detection algorithms are proposed.

The paper presents an approach for constructing a weighted bilingual dictionary of inflectional forms using as input data a traditional bilingual dictionary, and not parallel corpora. An algorithm is developed that generates all possible morphological (inflectional) forms and weights them using information on distribution of corresponding grammar sets (grammar information) in large corpora for each language. The algorithm also takes into account the compatibility of grammar sets in a language pair; for example, verb in past tense in language L normally is expected to be translated by verb in past tense in Language L'. We consider that the developed method is universal, i.e. can be applied to any pair of languages. The obtained dictionary is freely available. It can be used in several NLP tasks, for example, statistical machine translation.

Question Answering (QA) technology aims at providing relevant answers to natural language questions. Most Question Answering research has focused on mining document collections containing written texts to answer written questions. In addition to written sources, a large (and growing) amount of potentially interesting information appears in spoken documents, such as broadcast news, speeches, seminars, meetings or telephone conversations. The QAST track (Question-Answering on Speech Transcripts) was introduced in CLEF to investigate the problem of question answering in such audio documents. This paper describes in detail the evaluation protocol and tools designed and developed for the CLEF-QAST evaluation campaigns that have taken place between 2007 and 2009. We first remind the data, question sets, and submission procedures that were produced or set up during these three campaigns. As for the evaluation procedure, the interface that was developed to ease the assessors’ work is described. In addition, this paper introduces a methodology for a semi-automatic evaluation of QAST systems based on time slot comparisons. Finally, the QAST Evaluation Package 2007-2009 resulting from these evaluation campaigns is also introduced.

Subjectivity analysis and authorship attribution are very popular areas of research. However, work in these two areas has been done separately. We believe that by combining information about subjectivity in texts and authorship, the performance of both tasks can be improved. In the paper a personalized approach to opinion mining is presented, in which the notions of personal sense and idiolect are introduced; the approach is applied to the polarity classification task. It is assumed that different authors express their private states in text individually, and opinion mining results could be improved by analyzing texts by different authors separately. The hypothesis is tested on a corpus of movie reviews by ten authors. The results of applying the personalized approach to opinion mining are presented, confirming that the approach increases the performance of the opinion mining task. Automatic authorship attribution is further applied to model the personalized approach, classifying documents by their assumed authorship. Although the automatic authorship classification imposes a number of limitations on the dataset for further experiments, after overcoming these issues the authorship attribution technique modeling the personalized approach confirms the increase over the baseline with no authorship information used.

Research on automatic humor recognition has developed several features which discriminate funny text from ordinary text. The features have been demonstrated to work well when classifying the funniness of single sentences up to entire blogs. In this paper we focus on evaluating a set of the best humor features reported in the literature over a corpus retrieved from the Slashdot Web site. The corpus is categorized in a community-driven process according to the following tags: funny, informative, insightful, offtopic, flamebait, interesting and troll. These kinds of comments can be found on almost every large Web site; therefore, they impose a new challenge to humor retrieval since they come along with unique characteristics compared to other text types. If funny comments were retrieved accurately, they would be of a great entertainment value for the visitors of a given Web page. Our objective, thus, is to distinguish between an implicit funny comment from a not funny one. Our experiments are preliminary but nonetheless large-scale: 600,000 Web comments. We evaluate the classification accuracy of naive Bayes classifiers, decision trees, and support vector machines. The results suggested interesting findings.

WordNet has been used extensively as a resource for the Word Sense Disambiguation (WSD) task, both as a sense inventory and a repository of semantic relationships. Recently, we investigated the possibility to use it as a resource for the Geographical Information Retrieval task, more specifically for the toponym disambiguation task, which could be considered a specialization of WSD. We found that it would be very useful to assign to geographical entities inWordNet their coordinates, especially in order to implement geometric shapebased disambiguation methods. This paper presents Geo-WordNet, an automatic annotation of WordNet with geographical coordinates. The annotation has been carried out by extracting geographical synsets from WordNet, together with their holonyms and hypernyms, and comparing them to the entries in the Wikipedia-World geographical database. A weight was calculated for each of the candidate annotations, on the basis of matches found between the database entries and synset gloss, holonyms and hypernyms. The resulting resource may be used in Geographical Information Retrieval related tasks, especially for toponym disambiguation.

Although significant advances have been made recently in the Question Answering technology, more steps have to be undertaken in order to obtain better results. Moreover, the best systems at the CLEF and TREC evaluation exercises are very complex systems based on custom-built, expensive ontologies whose aim is to provide the systems with encyclopedic knowledge. In this paper we investigated the use of Wikipedia, the open domain encyclopedia, for the Question Answering task. Previous works considered Wikipedia as a resource where to look for the answers to the questions. We focused on some different aspects of the problem, such as the validation of the answers as returned by our Question Answering System and on the use of Wikipedia “categories” in order to determine a set of patterns that should fit with the expected answer. Validation consists in, given a possible answer, saying wether it is the right one or not. The possibility to exploit the categories ofWikipedia was not considered until now. We performed our experiments using the Spanish version of Wikipedia, with the set of questions of the last CLEF Spanish monolingual exercise. Results show that Wikipedia is a potentially useful resource for the Question Answering task.