Current automatic speech recognition (ASR) models are designed to be used across many languages and tasks without substantial changes. However, this broad language coverage hides performance gaps within languages, for example, across genders. Our study systematically evaluates the performance of two widely used multilingual ASR models on three datasets, encompassing 19 languages from eight language families and two speaking conditions. Our findings reveal clear gender disparities, with the advantaged group varying across languages and models. Surprisingly, those gaps are not explained by acoustic or lexical properties. However, probing internal model states reveals a correlation with gendered performance gap. That is, the easier it is to distinguish speaker gender in a language using probes, the more the gap reduces, favoring female speakers. Our results show that gender disparities persist even in state-of-the-art models. Our findings have implications for the improvement of multilingual ASR systems, underscoring the importance of accessibility to training data and nuanced evaluation to predict and mitigate gender gaps. We release all code and artifacts at https://github.com/g8a9/multilingual-asr-gender-gap.

Large Language Models (LLMs) exhibit remarkable text classification capabilities, excelling in zero- and few-shot learning (ZSL and FSL) scenarios. However, since they are trained on different datasets, performance varies widely across tasks between those models. Recent studies emphasize the importance of considering human label variation in data annotation. However, how this human label variation also applies to LLMs remains unexplored. Given this likely model specialization, we ask: Do aggregate LLM labels improve over individual models (as for human annotators)? We evaluate four recent instruction-tuned LLMs as “annotators” on five subjective tasks across four languages. We use ZSL and FSL setups and label aggregation from human annotation. Aggregations are indeed substantially better than any individual model, benefiting from specialization in diverse tasks or languages. Surprisingly, FSL does not surpass ZSL, as it depends on the quality of the selected examples. However, there seems to be no good information-theoretical strategy to select those. We find that no LLM method rivals even simple supervised models. We also discuss the tradeoffs in accuracy, cost, and moral/ethical considerations between LLM and human annotation.

The open-ended nature of language generation makes the evaluation of autoregressive large language models (LLMs) challenging. One common evaluation approach uses multiple-choice questions to limit the response space. The model is then evaluated by ranking the candidate answers by the log probability of the first token prediction. However, first-tokens may not consistently reflect the final response output, due to model’s diverse response styles such as starting with “Sure” or refusing to answer. Consequently, first-token evaluation is not indicative of model behaviour when interacting with users. But by how much? We evaluate how aligned first-token evaluation is with the text output along several dimensions, namely final option choice, refusal rate, choice distribution and robustness under prompt perturbation. Our results show that the two approaches are severely misaligned on all dimensions, reaching mismatch rates over 60%. Models heavily fine-tuned on conversational or safety data are especially impacted. Crucially, models remain misaligned even when we increasingly constrain prompts, i.e., force them to start with an option letter or example template. Our findings i) underscore the importance of inspecting the text output as well and ii) caution against relying solely on first-token evaluation.

Emotions play important epistemological and cognitive roles in our lives, revealing our values and guiding our actions. Previous work has shown that LLMs display biases in emotion attribution along gender lines. However, unlike gender, which says little about our values, religion, as a socio-cultural system, prescribes a set of beliefs and values for its followers. Religions, therefore, cultivate certain emotions. Moreover, these rules are explicitly laid out and interpreted by religious leaders. Using emotion attribution, we explore how different religions are represented in LLMs. We find that:Major religions in the US and European countries are represented with more nuance, displaying a more shaded model of their beliefs.Eastern religions like Hinduism and Buddhism are strongly stereotyped.Judaism and Islam are stigmatized – the models’ refusal skyrocket. We ascribe these to cultural bias in LLMs and the scarcity of NLP literature on religion. In the rare instances where religion is discussed, it is often in the context of toxic language, perpetuating the perception of these religions as inherently toxic. This finding underscores the urgent need to address and rectify these biases. Our research emphasizes the crucial role emotions play in shaping our lives and how our values influence them.

Without proper safeguards, large language models will readily follow malicious instructions and generate toxic content. This risk motivates safety efforts such as red-teaming and large-scale feedback learning, which aim to make models both helpful and harmless. However, there is a tension between these two objectives, since harmlessness requires models to refuse to comply with unsafe prompts, and thus not be helpful. Recent anecdotal evidence suggests that some models may have struck a poor balance, so that even clearly safe prompts are refused if they use similar language to unsafe prompts or mention sensitive topics. In this paper, we introduce a new test suite called XSTest to identify such eXaggerated Safety behaviours in a systematic way. XSTest comprises 250 safe prompts across ten prompt types that well-calibrated models should not refuse to comply with, and 200 unsafe prompts as contrasts that models, for most applications, should refuse. We describe XSTest’s creation and composition, and then use the test suite to highlight systematic failure modes in state-of-the-art language models as well as more general challenges in building safer language models.

Predictive models make mistakes and have biases. To combat both, we need to understand their predictions.Explainable AI (XAI) provides insights into models for vision, language, and tabular data. However, only a few approaches exist for speech classification models. Previous works focus on a selection of spoken language understanding (SLU) tasks, and most users find their explanations challenging to interpret.We propose a novel approach to explain speech classification models. It provides two types of insights. (i) Word-level. We measure the impact of each audio segment aligned with a word on the outcome. (ii) Paralinguistic. We evaluate how non-linguistic features (e.g., prosody and background noise) affect the outcome if perturbed.We validate our approach by explaining two state-of-the-art SLU models on two tasks in English and Italian. We test their plausibility with human subject ratings. Our results show that the explanations correctly represent the model’s inner workings and are plausible to humans.

Pre-trained language models consider the context of neighboring words and documents but lack any author context of the human generating the text. However, language depends on the author’s states, traits, social, situational, and environmental attributes, collectively referred to as human context (Soni et al., 2024). Human-centered natural language processing requires incorporating human context into language models. Currently, two methods exist: pre-training with 1) group-wise attributes (e.g., over-45-year-olds) or 2) individual traits. Group attributes are simple but coarse — not all 45-year-olds write the same way — while individual traits allow for more personalized representations, but require more complex modeling and data. It is unclear which approach benefits what tasks. We compare pre-training models with human context via 1) group attributes, 2) individual users, and 3) a combined approach on five user- and document-level tasks. Our results show that there is no best approach, but that human-centered language modeling holds avenues for different methods.

Using large language models (LLMs) for educational applications like dialogue-based teaching is a hot topic. Effective teaching, however, requires teachers to adapt the difficulty of content and explanations to the education level of their students. Even the best LLMs today struggle to do this well. If we want to improve LLMs on this adaptation task, we need to be able to measure adaptation success reliably. However, current Static metrics for text difficulty, like the Flesch-Kincaid Reading Ease score, are known to be crude and brittle. We, therefore, introduce and evaluate a new set of Prompt-based metrics for text difficulty. Based on a user study, we create Prompt-based metrics as inputs for LLMs. They leverage LLM’s general language understanding capabilities to capture more abstract and complex features than Static metrics. Regression experiments show that adding our Prompt-based metrics significantly improves text difficulty classification over Static metrics alone. Our results demonstrate the promise of using LLMs to evaluate text adaptation to different education levels.

Large language models (LLMs) reflect societal norms and biases, especially about gender. While societal biases and stereotypes have been extensively researched in various NLP applications, there is a surprising gap for emotion analysis. However, emotion and gender are closely linked in societal discourse. E.g., women are often thought of as more empathetic, while men’s anger is more socially accepted. To fill this gap, we present the first comprehensive study of gendered emotion attribution in five state-of-the-art LLMs (open- and closed-source). We investigate whether emotions are gendered, and whether these variations are based on societal stereotypes. We prompt the models to adopt a gendered persona and attribute emotions to an event like ‘When I had a serious argument with a dear person’. We then analyze the emotions generated by the models in relation to the gender-event pairs. We find that all models consistently exhibit gendered emotions, influenced by gender stereotypes. These findings are in line with established research in psychology and gender studies. Our study sheds light on the complex societal interplay between language, gender, and emotion. The reproduction of emotion stereotypes in LLMs allows us to use those models to study the topic in detail, but raises questions about the predictive use of those same LLMs for emotion applications.

The field of sociolinguistics has studied factors affecting language use for the last century. Labov (1964) and Bernstein (1960) showed that socioeconomic class strongly influences our accents, syntax and lexicon. However, despite growing concerns surrounding fairness and bias in Natural Language Processing (NLP), there is a dearth of studies delving into the effects it may have on NLP systems. We show empirically that NLP systems’ performance is affected by speakers’ SES, potentially disadvantaging less-privileged socioeconomic groups. We annotate a corpus of 95K utterances from movies with social class, ethnicity and geographical language variety and measure the performance of NLP systems on three tasks: language modelling, automatic speech recognition, and grammar error correction. We find significant performance disparities that can be attributed to socioeconomic status as well as ethnicity and geographical differences. With NLP technologies becoming ever more ubiquitous and quotidian, they must accommodate all language varieties to avoid disadvantaging already marginalised groups. We argue for the inclusion of socioeconomic class in future language technologies.

Much recent work seeks to evaluate values and opinions in large language models (LLMs) using multiple-choice surveys and questionnaires. Most of this work is motivated by concerns around real-world LLM applications. For example, politically-biased LLMs may subtly influence society when they are used by millions of people. Such real-world concerns, however, stand in stark contrast to the artificiality of current evaluations: real users do not typically ask LLMs survey questions. Motivated by this discrepancy, we challenge the prevailing *constrained* evaluation paradigm for values and opinions in LLMs and explore more realistic *unconstrained* evaluations. As a case study, we focus on the popular Political Compass Test (PCT). In a systematic review, we find that most prior work using the PCT *forces models to comply with the PCT’s multiple-choice format. We show that models give substantively different answers when not forced; that answers change depending on how models are forced; and that answers lack paraphrase robustness. Then, we demonstrate that models give different answers yet again in a more realistic open-ended answer setting. We distill these findings into recommendations and open challenges in evaluating values and opinions in LLMs.

Any report frames issues to favor a particular interpretation by highlighting or excluding certain aspects of a story. Despite the widespread use of framing in disinformation, framing properties and detection methods remain underexplored outside the English-speaking world. We explore how multilingual framing of the same issue differs systematically. We use eight years of Russia-backed disinformation campaigns, spanning 8k news articles in 4 languages targeting 15 countries. We find that disinformation campaigns consistently and intentionally favor specific framing, depending on the target language of the audience. We further discover how Russian-language articles consistently highlight selected frames depending on the region of the media coverage. We find that the two most prominent models for automatic frame analysis underperform and show high disagreement, highlighting the need for further research.

As diverse linguistic communities and users adopt Large Language Models (LLMs), assessing their safety across languages becomes critical. Despite ongoing efforts to align these models with safe and ethical guidelines, they can still be induced into unsafe behavior with jailbreaking, a technique in which models are prompted to act outside their operational guidelines. What research has been conducted on these vulnerabilities was predominantly on English, limiting the understanding of LLM behavior in other languages. We address this gap by investigating Many-Shot Jailbreaking (MSJ) in Italian, underscoring the importance of understanding LLM behavior in different languages. We base our analysis on a newly created Italian dataset to identify unique safety vulnerabilities in 4 families of open-source LLMs.We find that the models exhibit unsafe behaviors even with minimal exposure to harmful prompts, and–more alarmingly–this tendency rapidly escalates with more demonstrations.

Social scientists increasingly use demographically stratified social media data to study the attitudes, beliefs, and behavior of the general public. To facilitate such analyses, we construct, validate, and release publicly the representative DADIT dataset of 30M tweets of 20k Italian Twitter users, along with their bios and profile pictures. We enrich the user data with high-quality labels for gender, age, and location. DADIT enables us to train and compare the performance of various state-of-the-art models for the prediction of the gender and age of social media users. In particular, we investigate if tweets contain valuable information for the task, since popular classifiers like M3 don’t leverage them. Our best XLM-based classifier improves upon the commonly used competitor M3 by up to 53% F1. Especially for age prediction, classifiers profit from including tweets as features. We also confirm these findings on a German test set.

Emotions are a central aspect of communication. Consequently, emotion analysis (EA) is a rapidly growing field in natural language processing (NLP). However, there is no consensus on scope, direction, or methods. In this paper, we conduct a thorough review of 154 relevant NLP publications from the last decade. Based on this review, we address four different questions: (1) How are EA tasks defined in NLP? (2) What are the most prominent emotion frameworks and which emotions are modeled? (3) Is the subjectivity of emotions considered in terms of demographics and cultural factors? and (4) What are the primary NLP applications for EA? We take stock of trends in EA and tasks, emotion frameworks used, existing datasets, methods, and applications. We then discuss four lacunae: (1) the absence of demographic and cultural aspects does not account for the variation in how emotions are perceived, but instead assumes they are universally experienced in the same manner; (2) the poor fit of emotion categories from the two main emotion theories to the task; (3) the lack of standardized EA terminology hinders gap identification, comparison, and future goals; and (4) the absence of interdisciplinary research isolates EA from insights in other fields. Our work will enable more focused research into EA and a more holistic approach to modeling emotions in NLP.

Since Labov’s foundational 1964 work on the social stratification of language, linguistics has dedicated concerted efforts towards understanding the relationships between socio-demographic factors and language production and perception. Despite the large body of evidence identifying significant relationships between socio-demographic factors and language production, relatively few of these factors have been investigated in the context of NLP technology. While age and gender are well covered, Labov’s initial target, socio-economic class, is largely absent. We survey the existing Natural Language Processing (NLP) literature and find that only 20 papers even mention socio-economic status. However, the majority of those papers do not engage with class beyond collecting information of annotator-demographics. Given this research lacuna, we provide a definition of class that can be operationalised by NLP researchers, and argue for including socio-economic class in future language technologies.

We present the system proposed by the MilaNLP team for the Explainable Detection of Online Sexism (EDOS) shared task. We propose an ensemble modeling approach to combine different classifiers trained with domain adaptation objectives and standard fine-tuning. Our results show that the ensemble is more robust than individual models and that regularized models generate more “conservative” predictions, mitigating the effects of lexical overfitting.However, our error analysis also finds that many of the misclassified instances are debatable, raising questions about the objective annotatability of hate speech data.

As 3rd-person pronoun usage shifts to include novel forms, e.g., neopronouns, we need more research on identity-inclusive NLP. Exclusion is particularly harmful in one of the most popular NLP applications, machine translation (MT). Wrong pronoun translations can discriminate against marginalized groups, e.g., non-binary individuals (Dev et al., 2021). In this “reality check”, we study how three commercial MT systems translate 3rd-person pronouns. Concretely, we compare the translations of gendered vs. gender-neutral pronouns from English to five other languages (Danish, Farsi, French, German, Italian), and vice versa, from Danish to English.Our error analysis shows that the presence of a gender-neutral pronoun often leads to grammatical and semantic translation errors. Similarly, gender neutrality is often not preserved. By surveying the opinions of affected native speakers from diverse languages, we provide recommendations to address the issue in future MT research.

Many NLP tasks exhibit human label variation, where different annotators give different labels to the same texts. This variation is known to depend, at least in part, on the sociodemographics of annotators. Recent research aims to model individual annotator behaviour rather than predicting aggregated labels, and we would expect that sociodemographic information is useful for these models. On the other hand, the ecological fallacy states that aggregate group behaviour, such as the behaviour of the average female annotator, does not necessarily explain individual behaviour. To account for sociodemographics in models of individual annotator behaviour, we introduce group-specific layers to multi-annotator models. In a series of experiments for toxic content detection, we find that explicitly accounting for sociodemographic attributes in this way does not significantly improve model performance. This result shows that individual annotation behaviour depends on much more than just sociodemographics.

Demographic factors (e.g., gender or age) shape our language. Previous work showed that incorporating demographic factors can consistently improve performance for various NLP tasks with traditional NLP models. In this work, we investigate whether these previous findings still hold with state-of-the-art pretrained Transformer-based language models (PLMs). We use three common specialization methods proven effective for incorporating external knowledge into pretrained Transformers (e.g., domain-specific or geographic knowledge). We adapt the language representations for the demographic dimensions of gender and age, using continuous language modeling and dynamic multi-task learning for adaptation, where we couple language modeling objectives with the prediction of demographic classes. Our results, when employing a multilingual PLM, show substantial gains in task performance across four languages (English, German, French, and Danish), which is consistent with the results of previous work. However, controlling for confounding factors – primarily domain and language proficiency of Transformer-based PLMs – shows that downstream performance gains from our demographic adaptation do not actually stem from demographic knowledge. Our results indicate that demographic specialization of PLMs, while holding promise for positive societal impact, still represents an unsolved problem for (modern) NLP.

Work on hate speech has made considering rude and harmful examples in scientific publications inevitable. This situation raises various problems, such as whether or not to obscure profanities. While science must accurately disclose what it does, the unwarranted spread of hate speech can harm readers and increases its internet frequency. While maintaining publications’ professional appearance, obfuscating profanities makes it challenging to evaluate the content, especially for non-native speakers. Surveying 150 ACL papers, we discovered that obfuscation is usually used for English but not other languages, and even then, quite unevenly. We discuss the problems with obfuscation and suggest a multilingual community resource called PrOf with a Python module to standardize profanity obfuscation processes. We believe PrOf can help scientific publication policies to make hate speech work accessible and comparable, irrespective of language.

We report our efforts in identifying a set of previous human evaluations in NLP that would be suitable for a coordinated study examining what makes human evaluations in NLP more/less reproducible. We present our results and findings, which include that just 13% of papers had (i) sufficiently low barriers to reproduction, and (ii) enough obtainable information, to be considered for reproduction, and that all but one of the experiments we selected for reproduction was discovered to have flaws that made the meaningfulness of conducting a reproduction questionable. As a result, we had to change our coordinated study design from a reproduce approach to a standardise-then-reproduce-twice approach. Our overall (negative) finding that the great majority of human evaluations in NLP is not repeatable and/or not reproducible and/or too flawed to justify reproduction, paints a dire picture, but presents an opportunity for a rethink about how to design and report human evaluations in NLP.

Much work in natural language processing (NLP) relies on human annotation. The majority of this implicitly assumes that annotator’s labels are temporally stable, although the reality is that human judgements are rarely consistent over time. As a subjective annotation task, hate speech labels depend on annotator’s emotional and moral reactions to the language used to convey the message. Studies in Cognitive Science reveal a ‘foreign language effect’, whereby people take differing moral positions and perceive offensive phrases to be weaker in their second languages. Does this affect annotations as well? We conduct an experiment to investigate the impacts of (1) time and (2) different language conditions (English and German) on measurements of intra-annotator agreement in a hate speech labelling task. While we do not observe the expected lower stability in the different language condition, we find that overall agreement is significantly lower than is implicitly assumed in annotation tasks, which has important implications for dataset reproducibility in NLP.

Hate speech detection faces two significant challenges: 1) the limited availability of labeled data and 2) the high variability of hate speech across different contexts and languages. Prompting brings a ray of hope to these challenges. It allows injecting a model with task-specific knowledge without relying on labeled data. This paper explores zero-shot learning with prompting for hate speech detection. We investigate how well zero-shot learning can detect hate speech in 3 languages with limited labeled data. We experiment with various large language models and verbalizers on 8 benchmark datasets. Our findings highlight the impact of prompt selection on the results. They also suggest that prompting, specifically with recent large language models, can achieve performance comparable to and surpass fine-tuned models, making it a promising alternative for under-resourced languages. Our findings highlight the potential of prompting for hate speech detection and show how both the prompt and the model have a significant impact on achieving more accurate predictions in this task.

Labelled data is the foundation of most natural language processing tasks. However, labelling data is difficult and there often are diverse valid beliefs about what the correct data labels should be. So far, dataset creators have acknowledged annotator subjectivity, but rarely actively managed it in the annotation process. This has led to partly-subjective datasets that fail to serve a clear downstream use. To address this issue, we propose two contrasting paradigms for data annotation. The descriptive paradigm encourages annotator subjectivity, whereas the prescriptive paradigm discourages it. Descriptive annotation allows for the surveying and modelling of different beliefs, whereas prescriptive annotation enables the training of models that consistently apply one belief. We discuss benefits and challenges in implementing both paradigms, and argue that dataset creators should explicitly aim for one or the other to facilitate the intended use of their dataset. Lastly, we conduct an annotation experiment using hate speech data that illustrates the contrast between the two paradigms.

The social impact of natural language processing and its applications has received increasing attention. In this position paper, we focus on the problem of safety for end-to-end conversational AI. We survey the problem landscape therein, introducing a taxonomy of three observed phenomena: the Instigator, Yea-Sayer, and Impostor effects. We then empirically assess the extent to which current tools can measure these effects and current systems display them. We release these tools as part of a “first aid kit” (SafetyKit) to quickly assess apparent safety concerns. Our results show that, while current tools are able to provide an estimate of the relative safety of systems in various settings, they still have several shortcomings. We suggest several future directions and discuss ethical considerations.

Natural Language Processing (NLP) ‘s applied nature makes it necessary to select the most effective and robust models. Producing slightly higher performance is insufficient; we want to know whether this advantage will carry over to other data sets. Bootstrapped significance tests can indicate that ability. So while necessary, computing the significance of models’ performance differences has many levels of complexity. It can be tedious, especially when the experimental design has many conditions to compare and several runs of experiments. We present BooStSa, a tool that makes it easy to compute significance levels with the BOOtSTrap SAmpling procedure to evaluate models that predict not only standard hard labels but soft-labels (i.e., probability distributions over different classes) as well.

Current language technology is ubiquitous and directly influences individuals’ lives worldwide. Given the recent trend in AI on training and constantly releasing new and powerful large language models (LLMs), there is a need to assess their biases and potential concrete consequences. While some studies have highlighted the shortcomings of these models, there is only little on the negative impact of LLMs on LGBTQIA+ individuals. In this paper, we investigated a state-of-the-art template-based approach for measuring the harmfulness of English LLMs sentence completion when the subjects belong to the LGBTQIA+ community. Our findings show that, on average, the most likely LLM-generated completion is an identity attack 13% of the time. Our results raise serious concerns about the applicability of these models in production environments.

Over the last several years, end-to-end neural conversational agents have vastly improved their ability to carry unrestricted, open-domain conversations with humans. However, these models are often trained on large datasets from the Internet and, as a result, may learn undesirable behaviours from this data, such as toxic or otherwise harmful language. Thus, researchers must wrestle with how and when to release these models. In this paper, we survey recent and related work to highlight tensions between values, potential positive impact, and potential harms. We also provide a framework to support practitioners in deciding whether and how to release these models, following the tenets of value-sensitive design.

Hate speech is a global phenomenon, but most hate speech datasets so far focus on English-language content. This hinders the development of more effective hate speech detection models in hundreds of languages spoken by billions across the world. More data is needed, but annotating hateful content is expensive, time-consuming and potentially harmful to annotators. To mitigate these issues, we explore data-efficient strategies for expanding hate speech detection into under-resourced languages. In a series of experiments with mono- and multilingual models across five non-English languages, we find that 1) a small amount of target-language fine-tuning data is needed to achieve strong performance, 2) the benefits of using more such data decrease exponentially, and 3) initial fine-tuning on readily-available English data can partially substitute target-language data and improve model generalisability. Based on these findings, we formulate actionable recommendations for hate speech detection in low-resource language settings.

Fairness and environmental impact are important research directions for the sustainable development of artificial intelligence. However, while each topic is an active research area in natural language processing (NLP), there is a surprising lack of research on the interplay between the two fields. This lacuna is highly problematic, since there is increasing evidence that an exclusive focus on fairness can actually hinder environmental sustainability, and vice versa. In this work, we shed light on this crucial intersection in NLP by (1) investigating the efficiency of current fairness approaches through surveying example methods for reducing unfair stereotypical bias from the literature, and (2) evaluating a common technique to reduce energy consumption (and thus environmental impact) of English NLP models, knowledge distillation (KD), for its impact on fairness. In this case study, we evaluate the effect of important KD factors, including layer and dimensionality reduction, with respect to: (a) performance on the distillation task (natural language inference and semantic similarity prediction), and (b) multiple measures and dimensions of stereotypical bias (e.g., gender bias measured via the Word Embedding Association Test). Our results lead us to clarify current assumptions regarding the effect of KD on unfair bias: contrary to other findings, we show that KD can actually decrease model fairness.

Pre-trained language models (PLMs) have outperformed other NLP models on a wide range of tasks. Opting for a more thorough understanding of their capabilities and inner workings, researchers have established the extend to which they capture lower-level knowledge like grammaticality, and mid-level semantic knowledge like factual understanding. However, there is still little understanding of their knowledge of higher-level aspects of language. In particular, despite the importance of sociodemographic aspects in shaping our language, the questions of whether, where, and how PLMs encode these aspects, e.g., gender or age, is still unexplored. We address this research gap by probing the sociodemographic knowledge of different single-GPU PLMs on multiple English data sets via traditional classifier probing and information-theoretic minimum description length probing. Our results show that PLMs do encode these sociodemographics, and that this knowledge is sometimes spread across the layers of some of the tested PLMs. We further conduct a multilingual analysis and investigate the effect of supplementary training to further explore to what extent, where, and with what amount of pre-training data the knowledge is encoded. Our overall results indicate that sociodemographic knowledge is still a major challenge for NLP. PLMs require large amounts of pre-training data to acquire the knowledge and models that excel in general language understanding do not seem to own more knowledge about these aspects.

Well-annotated data is a prerequisite for good Natural Language Processing models. Too often, though, annotation decisions are governed by optimizing time or annotator agreement. We make a case for nuanced efforts in an interdisciplinary setting for annotating offensive online speech. Detecting offensive content is rapidly becoming one of the most important real-world NLP tasks. However, most datasets use a single binary label, e.g., for hate or incivility, even though each concept is multi-faceted. This modeling choice severely limits nuanced insights, but also performance.We show that a more fine-grained multi-label approach to predicting incivility and hateful or intolerant content addresses both conceptual and performance issues.We release a novel dataset of over 40,000 tweets about immigration from the US and UK, annotated with six labels for different aspects of incivility and intolerance.Our dataset not only allows for a more nuanced understanding of harmful speech online, models trained on it also outperform or match performance on benchmark datasets

Twitter data have become essential to Natural Language Processing (NLP) and social science research, driving various scientific discoveries in recent years. However, the textual data alone are often not enough to conduct studies: especially, social scientists need more variables to perform their analysis and control for various factors. How we augment this information, such as users’ location, age, or tweet sentiment, has ramifications for anonymity and reproducibility, and requires dedicated effort. This paper describes Twitter-Demographer, a simple, flow-based tool to enrich Twitter data with additional information about tweets and users. \tool is aimed at NLP practitioners, psycho-linguists, and (computational) social scientists who want to enrich their datasets with aggregated information, facilitating reproducibility, and providing algorithmic privacy-by-design measures for pseudo-anonymity. We discuss our design choices, inspired by the flow-based programming paradigm, to use black-box components that can easily be chained together and extended. We also analyze the ethical issues related to the use of this tool, and the built-in measures to facilitate pseudo-anonymity.

Natural Language Processing (NLP) models risk overfitting to specific terms in the training data, thereby reducing their performance, fairness, and generalizability. E.g., neural hate speech detection models are strongly influenced by identity terms like gay, or women, resulting in false positives, severe unintended bias, and lower performance. Most mitigation techniques use lists of identity terms or samples from the target domain during training. However, this approach requires a-priori knowledge and introduces further bias if important terms are neglected. Instead, we propose a knowledge-free Entropy-based Attention Regularization (EAR) to discourage overfitting to training-specific terms. An additional objective function penalizes tokens with low self-attention entropy. We fine-tune BERT via EAR: the resulting model matches or exceeds state-of-the-art performance for hate speech classification and bias metrics on three benchmark corpora in English and Italian.EAR also reveals overfitting terms, i.e., terms most likely to induce bias, to help identify their effect on the model, task, and predictions.

Meaning is context-dependent, but many properties of language (should) remain the same even if we transform the context. For example, sentiment or speaker properties should be the same in a translation and original of a text. We introduce language invariant properties: i.e., properties that should not change when we transform text, and how they can be used to quantitatively evaluate the robustness of transformation algorithms. Language invariant properties can be used to define novel benchmarks to evaluate text transformation methods. In our work we use translation and paraphrasing as examples, but our findings apply more broadly to any transformation. Our results indicate that many NLP transformations change properties. We additionally release a tool as a proof of concept to evaluate the invariance of transformation applications.

Transformer-based Natural Language Processing models have become the standard for hate speech detection. However, the unconscious use of these techniques for such a critical task comes with negative consequences. Various works have demonstrated that hate speech classifiers are biased. These findings have prompted efforts to explain classifiers, mainly using attribution methods. In this paper, we provide the first benchmark study of interpretability approaches for hate speech detection. We cover four post-hoc token attribution approaches to explain the predictions of Transformer-based misogyny classifiers in English and Italian. Further, we compare generated attributions to attention analysis. We find that only two algorithms provide faithful explanations aligned with human expectations. Gradient-based methods and attention, however, show inconsistent outputs, making their value for explanations questionable for hate speech detection tasks.

Detecting emotion in text allows social and computational scientists to study how people behave and react to online events. However, developing these tools for different languages requires data that is not always available. This paper collects the available emotion detection datasets across 19 languages. We train a multilingual emotion prediction model for social media data, XLM-EMO. The model shows competitive performance in a zero-shot setting, suggesting it is helpful in the context of low-resource languages. We release our model to the community so that interested researchers can directly use it.

The maturity level of language models is now at a stage in which many companies rely on them to solve various tasks. However, while research has shown how biased and harmful these models are, systematic ways of integrating social bias tests into development pipelines are still lacking. This short paper suggests how to use these verification techniques in development pipelines. We take inspiration from software testing and suggest addressing social bias evaluation as software testing. We hope to open a discussion on the best methodologies to handle social bias testing in language models.

The world of pronouns is changing – from a closed word class with few members to an open set of terms to reflect identities. However, Natural Language Processing (NLP) barely reflects this linguistic shift, resulting in the possible exclusion of non-binary users, even though recent work outlined the harms of gender-exclusive language technology. The current modeling of 3rd person pronouns is particularly problematic. It largely ignores various phenomena like neopronouns, i.e., novel pronoun sets that are not (yet) widely established. This omission contributes to the discrimination of marginalized and underrepresented groups, e.g., non-binary individuals. It thus prevents gender equality, one of the UN’s sustainable development goals (goal 5). Further, other identity-expressions beyond gender are ignored by current NLP technology. This paper provides an overview of 3rd person pronoun issues for NLP. Based on our observations and ethical considerations, we define a series of five desiderata for modeling pronouns in language technology, which we validate through a survey. We evaluate existing and novel modeling approaches w.r.t. these desiderata qualitatively and quantify the impact of a more discrimination-free approach on an established benchmark dataset.

Topic models extract groups of words from documents, whose interpretation as a topic hopefully allows for a better understanding of the data. However, the resulting word groups are often not coherent, making them harder to interpret. Recently, neural topic models have shown improvements in overall coherence. Concurrently, contextual embeddings have advanced the state of the art of neural models in general. In this paper, we combine contextualized representations with neural topic models. We find that our approach produces more meaningful and coherent topics than traditional bag-of-words topic models and recent neural models. Our results indicate that future improvements in language models will translate into better topic models.

Natural language processing (NLP) applications are now more powerful and ubiquitous than ever before. With rapidly developing (neural) models and ever-more available data, current NLP models have access to more information than any human speaker during their life. Still, it would be hard to argue that NLP models have reached human-level capacity. In this position paper, we argue that the reason for the current limitations is a focus on information content while ignoring language’s social factors. We show that current NLP systems systematically break down when faced with interpreting the social factors of language. This limits applications to a subset of information-related tasks and prevents NLP from reaching human-level performance. At the same time, systems that incorporate even a minimum of social factors already show remarkable improvements. We formalize a taxonomy of seven social factors based on linguistic theory and exemplify current failures and emerging successes for each of them. We suggest that the NLP community address social factors to get closer to the goal of human-like language understanding.

Language models have revolutionized the field of NLP. However, language models capture and proliferate hurtful stereotypes, especially in text generation. Our results show that 4.3% of the time, language models complete a sentence with a hurtful word. These cases are not random, but follow language and gender-specific patterns. We propose a score to measure hurtful sentence completions in language models (HONEST). It uses a systematic template- and lexicon-based bias evaluation methodology for six languages. Our findings suggest that these models replicate and amplify deep-seated societal stereotypes about gender roles. Sentence completions refer to sexual promiscuity when the target is female in 9% of the time, and in 4% to homosexuality when the target is male. The results raise questions about the use of these models in production settings.

Supervised learning assumes that a ground truth label exists. However, the reliability of this ground truth depends on human annotators, who often disagree. Prior work has shown that this disagreement can be helpful in training models. We propose a novel method to incorporate this disagreement as information: in addition to the standard error computation, we use soft-labels (i.e., probability distributions over the annotator labels) as an auxiliary task in a multi-task neural network. We measure the divergence between the predictions and the target soft-labels with several loss-functions and evaluate the models on various NLP tasks. We find that the soft-label prediction auxiliary task reduces the penalty for errors on ambiguous entities, and thereby mitigates overfitting. It significantly improves performance across tasks, beyond the standard approach and prior work.

While emotions are universal aspects of human psychology, they are expressed differently across different languages and cultures. We introduce a new data set of over 530k anonymized public Facebook posts across 18 languages, labeled with five different emotions. Using multilingual BERT embeddings, we show that emotions can be reliably inferred both within and across languages. Zero-shot learning produces promising results for low-resource languages. Following established theories of basic emotions, we provide a detailed analysis of the possibilities and limits of cross-lingual emotion classification. We find that structural and typological similarity between languages facilitates cross-lingual learning, as well as linguistic diversity of training data. Our results suggest that there are commonalities underlying the expression of emotion in different languages. We publicly release the anonymized data for future research.

While sentiment analysis is a popular task to understand people’s reactions online, we often need more nuanced information: is the post negative because the user is angry or sad? An abundance of approaches have been introduced for tackling these tasks, also for Italian, but they all treat only one of the tasks. We introduce FEEL-IT, a novel benchmark corpus of Italian Twitter posts annotated with four basic emotions: anger, fear, joy, sadness. By collapsing them, we can also do sentiment analysis. We evaluate our corpus on benchmark datasets for both emotion and sentiment classification, obtaining competitive results. We release an open-source Python library, so researchers can use a model trained on FEEL-IT for inferring both sentiments and emotions from Italian text.

The paper describes the MilaNLP team’s submission (Bocconi University, Milan) in the WASSA 2021 Shared Task on Empathy Detection and Emotion Classification. We focus on Track 2 - Emotion Classification - which consists of predicting the emotion of reactions to English news stories at the essay-level. We test different models based on multi-task and multi-input frameworks. The goal was to better exploit all the correlated information given in the data set. We find, though, that empathy as an auxiliary task in multi-task learning and demographic attributes as additional input provide worse performance with respect to single-task learning. While the result is competitive in terms of the competition, our results suggest that emotion and empathy are not related tasks - at least for the purpose of prediction.

Many data sets (e.g., reviews, forums, news, etc.) exist parallelly in multiple languages. They all cover the same content, but the linguistic differences make it impossible to use traditional, bag-of-word-based topic models. Models have to be either single-language or suffer from a huge, but extremely sparse vocabulary. Both issues can be addressed by transfer learning. In this paper, we introduce a zero-shot cross-lingual topic model. Our model learns topics on one language (here, English), and predicts them for unseen documents in different languages (here, Italian, French, German, and Portuguese). We evaluate the quality of the topic predictions for the same document in different languages. Our results show that the transferred topics are coherent and stable across languages, which suggests exciting future research directions.

Spotting a lie is challenging but has an enormous potential impact on security as well as private and public safety. Several NLP methods have been proposed to classify texts as truthful or deceptive. In most cases, however, the target texts’ preceding context is not considered. This is a severe limitation, as any communication takes place in context, not in a vacuum, and context can help to detect deception. We study a corpus of Italian dialogues containing deceptive statements and implement deep neural models that incorporate various linguistic contexts. We establish a new state-of-the-art identifying deception and find that not all context is equally useful to the task. Only the texts closest to the target, if from the same speaker (rather than questions by an interlocutor), boost performance. We also find that the semantic information in language models such as BERT contributes to the performance. However, BERT alone does not capture the implicit knowledge of deception cues: its contribution is conditional on the concurrent use of attention to learn cues from BERT’s representations.

Evaluation is of paramount importance in data-driven research fields such as Natural Language Processing (NLP) and Computer Vision (CV). Current evaluation practice largely hinges on the existence of a single “ground truth” against which we can meaningfully compare the prediction of a model. However, this comparison is flawed for two reasons. 1) In many cases, more than one answer is correct. 2) Even where there is a single answer, disagreement among annotators is ubiquitous, making it difficult to decide on a gold standard. We argue that the current methods of adjudication, agreement, and evaluation need serious reconsideration. Some researchers now propose to minimize disagreement and to fix datasets. We argue that this is a gross oversimplification, and likely to conceal the underlying complexity. Instead, we suggest that we need to better capture the sources of disagreement to improve today’s evaluation practice. We discuss three sources of disagreement: from the annotator, the data, and the context, and show how this affects even seemingly objective tasks. Datasets with multiple annotations are becoming more common, as are methods to integrate disagreement into modeling. The logical next step is to extend this to evaluation.

This paper presents the results of the VarDial Evaluation Campaign 2020 organized as part of the seventh workshop on Natural Language Processing (NLP) for Similar Languages, Varieties and Dialects (VarDial), co-located with COLING 2020. The campaign included three shared tasks each focusing on a different challenge of language and dialect identification: Romanian Dialect Identification (RDI), Social Media Variety Geolocation (SMG), and Uralic Language Identification (ULI). The campaign attracted 30 teams who enrolled to participate in one or multiple shared tasks and 14 of them submitted runs across the three shared tasks. Finally, 11 papers describing participating systems are published in the VarDial proceedings and referred to in this report.

The main goal of machine translation has been to convey the correct content. Stylistic considerations have been at best secondary. We show that as a consequence, the output of three commercial machine translation systems (Bing, DeepL, Google) make demographically diverse samples from five languages “sound” older and more male than the original. Our findings suggest that translation models reflect demographic bias in the training data. This opens up interesting new research avenues in machine translation to take stylistic considerations into account.

An increasing number of natural language processing papers address the effect of bias on predictions, introducing mitigation techniques at different parts of the standard NLP pipeline (data and models). However, these works have been conducted individually, without a unifying framework to organize efforts within the field. This situation leads to repetitive approaches, and focuses overly on bias symptoms/effects, rather than on their origins, which could limit the development of effective countermeasures. In this paper, we propose a unifying predictive bias framework for NLP. We summarize the NLP literature and suggest general mathematical definitions of predictive bias. We differentiate two consequences of bias: outcome disparities and error disparities, as well as four potential origins of biases: label bias, selection bias, model overamplification, and semantic bias. Our framework serves as an overview of predictive bias in NLP, integrating existing work into a single structure, and providing a conceptual baseline for improved frameworks.

To raise awareness among future NLP practitioners and prevent inertia in the field, we need to place ethics in the curriculum for all NLP students—not as an elective, but as a core part of their education. Our goal in this tutorial is to empower NLP researchers and practitioners with tools and resources to teach others about how to ethically apply NLP techniques. We will present both high-level strategies for developing an ethics-oriented curriculum, based on experience and best practices, as well as specific sample exercises that can be brought to a classroom. This highly interactive work session will culminate in a shared online resource page that pools lesson plans, assignments, exercise ideas, reading suggestions, and ideas from the attendees. Though the tutorial will focus particularly on examples for university classrooms, we believe these ideas can extend to company-internal workshops or tutorials in a variety of organizations. In this setting, a key lesson is that there is no single approach to ethical NLP: each project requires thoughtful consideration about what steps can be taken to best support people affected by that project. However, we can learn (and teach) what issues to be aware of, what questions to ask, and what strategies are available to mitigate harm.

When interacting with each other, we motivate, advise, inform, show love or power towards our peers. However, the way we interact may also hold some indication on how successful we are, as people often try to help each other to achieve their goals. We study the chat interactions of thousands of aspiring entrepreneurs who discuss and develop business models. We manually annotate a set of about 5,500 chat interactions with four dimensions of interaction styles (motivation, cooperation, equality, advice). We find that these styles can be reliably predicted, and that the communication styles can be used to predict a number of indices of business success. Our findings indicate that successful communicators are also successful in other domains.

Several linguistic studies have shown the prevalence of various lexical and grammatical patterns in texts authored by a person of a particular gender, but models for part-of-speech tagging and dependency parsing have still not adapted to account for these differences. To address this, we annotate the Wall Street Journal part of the Penn Treebank with the gender information of the articles’ authors, and build taggers and parsers trained on this data that show performance differences in text written by men and women. Further analyses reveal numerous part-of-speech tags and syntactic relations whose prediction performances benefit from the prevalence of a specific gender in the training data. The results underscore the importance of accounting for gendered differences in syntactic tasks, and outline future venues for developing more accurate taggers and parsers. We release our data to the research community.

User reviews provide a significant source of information for companies to understand their market and audience. In order to discover broad trends in this source, researchers have typically used topic models such as Latent Dirichlet Allocation (LDA). However, while there are metrics to choose the “best” number of topics, it is not clear whether the resulting topics can also provide in-depth, actionable product analysis. Our paper examines this issue by analyzing user reviews from the Best Buy US website for smart speakers. Using coherence scores to choose topics, we test whether the results help us to understand user interests and concerns. We find that while coherence scores are a good starting point to identify a number of topics, it still requires manual adaptation based on domain knowledge to provide market insights. We show that the resulting dimensions capture brand performance and differences, and differentiate the market into two distinct groups with different properties.

Geolocation, predicting the location of a post based on text and other information, has a huge potential for several social media applications. Typically, the problem is modeled as either multi-class classification or regression. In the first case, the classes are geographic areas previously identified; in the second, the models directly predict geographic coordinates. The former requires discretization of the coordinates, but yields better performance. The latter is potentially more precise and true to the nature of the problem, but often results in worse performance. We propose to combine the two approaches in an attentionbased multitask convolutional neural network that jointly predicts both discrete locations and continuous geographic coordinates. We evaluate the multi-task (MTL) model against singletask models and prior work. We find that MTL significantly improves performance, reporting large gains on one data set, but also note that the correlation between labels and coordinates has a marked impact on the effectiveness of including a regression task.

Prior research has shown that geolocation can be substantially improved by including user network information. While effective, it suffers from the curse of dimensionality, since networks are usually represented as sparse adjacency matrices of connections, which grow exponentially with the number of users. In order to incorporate this information, we therefore need to limit the network size, in turn limiting performance and risking sample bias. In this paper, we address these limitations by instead using dense network representations. We explore two methods to learn continuous node representations from either 1) the network structure with node2vec (Grover and Leskovec, 2016), or 2) textual user mentions via doc2vec (Le and Mikolov, 2014). We combine both methods with input from social media posts in an attention-based convolutional neural network and evaluate the contribution of each component on geolocation performance. Our method enables us to incorporate arbitrarily large networks in a fixed-length vector, without limiting the network size. Our models achieve competitive results with similar state-of-the-art methods, but with much fewer model parameters, while being applicable to networks of virtually any size.

Geolocating social media posts relies on the assumption that language carries sufficient geographic information. However, locations are usually given as continuous latitude/longitude tuples, so we first need to define discrete geographic regions that can serve as labels. Most studies use some form of clustering to discretize the continuous coordinates (Han et al., 2016). However, the resulting regions do not always correspond to existing linguistic areas. Consequently, accuracy at 100 miles tends to be good, but degrades for finer-grained distinctions, when different linguistic regions get lumped together. We describe a new algorithm, Point-to-City (P2C), an iterative k-d tree-based method for clustering geographic coordinates and associating them with towns. We create three sets of labels at different levels of granularity, and compare performance of a state-of-the-art geolocation model trained and tested with P2C labels to one with regular k-d tree labels. Even though P2C results in substantially more labels than the baseline, model accuracy increases significantly over using traditional labels at the fine-grained level, while staying comparable at 100 miles. The results suggest that identifying meaningful linguistic areas is crucial for improving geolocation at a fine-grained level.

The analysis of crowdsourced annotations in natural language processing is concerned with identifying (1) gold standard labels, (2) annotator accuracies and biases, and (3) item difficulties and error patterns. Traditionally, majority voting was used for 1, and coefficients of agreement for 2 and 3. Lately, model-based analysis of corpus annotations have proven better at all three tasks. But there has been relatively little work comparing them on the same datasets. This paper aims to fill this gap by analyzing six models of annotation, covering different approaches to annotator ability, item difficulty, and parameter pooling (tying) across annotators and items. We evaluate these models along four aspects: comparison to gold labels, predictive accuracy for new annotations, annotator characterization, and item difficulty, using four datasets with varying degrees of noise in the form of random (spammy) annotators. We conclude with guidelines for model selection, application, and implementation.

Over the years, natural language processing has increasingly focused on tasks that can be solved by statistical models, but ignored the social aspects of language. These limitations are in large part due to historically available data and the limitations of the models, but have narrowed our focus and biased the tools demographically. However, with the increased availability of data sets including socio-demographic information and more expressive (neural) models, we have the opportunity to address both issues. I argue that this combination can broaden the focus of NLP to solve a whole new range of tasks, enable us to generate novel linguistic insights, and provide fairer tools for everyone.

Newspapers need to attract readers with headlines, anticipating their readers’ preferences. These preferences rely on topical, structural, and lexical factors. We model each of these factors in a multi-task GRU network to predict headline popularity. We find that pre-trained word embeddings provide significant improvements over untrained embeddings, as do the combination of two auxiliary tasks, news-section prediction and part-of-speech tagging. However, we also find that performance is very similar to that of a simple Logistic Regression model over character n-grams. Feature analysis reveals structural patterns of headline popularity, including the use of forward-looking deictic expressions and second person pronouns.

Most text-classification approaches represent the input based on textual features, either feature-based or continuous. However, this ignores strong non-linguistic similarities like homophily: people within a demographic group use language more similar to each other than to non-group members. We use homophily cues to retrofit text-based author representations with non-linguistic information, and introduce a trade-off parameter. This approach increases in-class similarity between authors, and improves classification performance by making classes more linearly separable. We evaluate the effect of our method on two author-attribute prediction tasks with various training-set sizes and parameter settings. We find that our method can significantly improve classification performance, especially when the number of labels is large and limited labeled data is available. It is potentially applicable as preprocessing step to any text-classification task.

Dialects are one of the main drivers of language variation, a major challenge for natural language processing tools. In most languages, dialects exist along a continuum, and are commonly discretized by combining the extent of several preselected linguistic variables. However, the selection of these variables is theory-driven and itself insensitive to change. We use Doc2Vec on a corpus of 16.8M anonymous online posts in the German-speaking area to learn continuous document representations of cities. These representations capture continuous regional linguistic distinctions, and can serve as input to downstream NLP tasks sensitive to regional variation. By incorporating geographic information via retrofitting and agglomerative clustering with structure, we recover dialect areas at various levels of granularity. Evaluating these clusters against an existing dialect map, we achieve a match of up to 0.77 V-score (harmonic mean of cluster completeness and homogeneity). Our results show that representation learning with retrofitting offers a robust general method to automatically expose dialectal differences and regional variation at a finer granularity than was previously possible.

Language contains information about the author’s demographic attributes as well as their mental state, and has been successfully leveraged in NLP to predict either one alone. However, demographic attributes and mental states also interact with each other, and we are the first to demonstrate how to use them jointly to improve the prediction of mental health conditions across the board. We model the different conditions as tasks in a multitask learning (MTL) framework, and establish for the first time the potential of deep learning in the prediction of mental health from online user-generated text. The framework we propose significantly improves over all baselines and single-task models for predicting mental health conditions, with particularly significant gains for conditions with limited data. In addition, our best MTL model can predict the presence of conditions (neuroatypicality) more generally, further reducing the error of the strong feed-forward baseline.

Geolocation is the task of identifying a social media user’s primary location, and in natural language processing, there is a growing literature on to what extent automated analysis of social media posts can help. However, not all content features are equally revealing of a user’s location. In this paper, we evaluate nine name entity (NE) types. Using various metrics, we find that GEO-LOC, FACILITY and SPORT-TEAM are more informative for geolocation than other NE types. Using these types, we improve geolocation accuracy and reduce distance error over various famous text-based methods.

Most state-of-the-art information extraction approaches rely on token-level labels to find the areas of interest in text. Unfortunately, these labels are time-consuming and costly to create, and consequently, not available for many real-life IE tasks. To make matters worse, token-level labels are usually not the desired output, but just an intermediary step. End-to-end (E2E) models, which take raw text as input and produce the desired output directly, need not depend on token-level labels. We propose an E2E model based on pointer networks, which can be trained directly on pairs of raw input and output text. We evaluate our model on the ATIS data set, MIT restaurant corpus and the MIT movie corpus and compare to neural baselines that do use token-level labels. We achieve competitive results, within a few percentage points of the baselines, showing the feasibility of E2E information extraction without the need for token-level labels. This opens up new possibilities, as for many tasks currently addressed by human extractors, raw input and output data are available, but not token-level labels.

Language varies not only between countries, but also along regional and socio-demographic lines. This variation is one of the driving factors behind language change. However, investigating language variation is a complex undertaking: the more factors we want to consider, the more data we need. Traditional qualitative methods are not well-suited to do this, an therefore restricted to isolated factors. This reduction limits the potential insights, and risks attributing undue importance to easily observed factors. While there is a large interest in linguistics to increase the quantitative aspect of such studies, it requires training in both variational linguistics and computational methods, a combination that is still not common. We take a first step here to alleviating the problem by providing an interface, www.languagevariation.com, to explore large-scale language variation along multiple socio-demographic factors – without programming knowledge. It makes use of large amounts of data and provides statistical analyses, maps, and interactive features that will enable scholars to explore language variation in a data-driven way.

We develop a supersense taxonomy for adjectives, based on that of GermaNet, and apply it to English adjectives in WordNet using human annotation and supervised classification. Results show that accuracy for automatic adjective type classification is high, but synsets are considerably more difficult to classify, even for trained human annotators. We release the manually annotated data, the classifier, and the induced supersense labeling of 12,304 WordNet adjective synsets.

We present two new NER datasets for Twitter; a manually annotated set of 1,467 tweets (kappa=0.942) and a set of 2,975 expert-corrected, crowdsourced NER annotated tweets from the dataset described in Finin et al. (2010). In our experiments with these datasets, we observe two important points: (a) language drift on Twitter is significant, and while off-the-shelf systems have been reported to perform well on in-sample data, they often perform poorly on new samples of tweets, (b) state-of-the-art performance across various datasets can be obtained from crowdsourced annotations, making it more feasible to “catch up” with language drift.

Several works in Natural Language Processing have recently looked into part-of-speech annotation of Twitter data and typically used their own data sets. Since conventions on Twitter change rapidly, models often show sample bias. Training on a combination of the existing data sets should help overcome this bias and produce more robust models than any trained on the individual corpora. Unfortunately, combining the existing corpora proves difficult: many of the corpora use proprietary tag sets that have little or no overlap. Even when mapped to a common tag set, the different corpora systematically differ in their treatment of various tags and tokens. This includes both pre-processing decisions, as well as default labels for frequent tokens, thus exhibiting data bias and label bias, respectively. Only if we address these biases can we combine the existing data sets to also overcome sample bias. We present a systematic study of several Twitter POS data sets, the problems of label and data bias, discuss their effects on model performance, and show how to overcome them to learn models that perform well on various test sets, achieving relative error reduction of up to 21%.