While social media offer great communication opportunities, they also increase the vulnerability of young people to threatening situations online. Recent studies report that cyberbullying constitutes a growing problem among youngsters. Successful prevention depends on the adequate detection of potentially harmful messages and the information overload on the Web requires intelligent systems to identify potential risks automatically. The focus of this paper is on automatic cyberbullying detection in social media text by modelling posts written by bullies, victims, and bystanders of online bullying. We describe the collection and fine-grained annotation of a cyberbullying corpus for English and Dutch and perform a series of binary classification experiments to determine the feasibility of automatic cyberbullying detection. We make use of linear support vector machines exploiting a rich feature set and investigate which information sources contribute the most for the task. Experiments on a hold-out test set reveal promising results for the detection of cyberbullying-related posts. After optimisation of the hyperparameters, the classifier yields an F1 score of 64% and 61% for English and Dutch respectively, and considerably outperforms baseline systems.

The detection of online cyberbullying has seen an increase in societal importance, popularity in research, and available open data. Nevertheless, while computational power and affordability of resources continue to increase, the access restrictions on high-quality data limit the applicability of state-of-the-art techniques. Consequently, much of the recent research uses small, heterogeneous datasets, without a thorough evaluation of applicability. In this paper, we further illustrate these issues, as we (i) evaluate many publicly available resources for this task and demonstrate difficulties with data collection. These predominantly yield small datasets that fail to capture the required complex social dynamics and impede direct comparison of progress. We (ii) conduct an extensive set of experiments that indicate a general lack of cross-domain generalization of classifiers trained on these sources, and openly provide this framework to replicate and extend our evaluation criteria. Finally, we (iii) present an effective crowdsourcing method: simulating real-life bullying scenarios in a lab setting generates plausible data that can be effectively used to enrich real data. This largely circumvents the restrictions on data that can be collected, and increases classifier performance. We believe these contributions can aid in improving the empirical practices of future research in the field.

We present results of the first gender classification experiments on Slovene text to our knowledge. Inspired by the TwiSty corpus and experiments , we employed the Janes corpus and its gender annotations to perform gender classification experiments on Twitter text comparing a token-based and a lemma-based approach. We find that the token-based approach (92.6% accuracy), containing gender markings related to the author, outperforms the lemma-based approach by about 5%. Especially in the lemmatized version, we also observe stylistic and contentbased differences in writing between men (e.g., more profane language, numerals and beer mentions) and women (e.g., more pronouns, emoticons and character flooding). Many of our findings corroborate previous research on other languages.

Compounding, the process of combining several simplex words into a complex whole, is a productive process in a wide range of languages. In particular, concatenative compounding, in which the components are "glued" together, leads to problems, for instance, in computational tools that rely on a predefined lexicon. Here we present the AuCoPro project, which focuses on compounding in the closely related languages Afrikaans and Dutch. The project consists of subprojects focusing on compound splitting (identifying the boundaries of the components) and compound semantics (identifying semantic relations between the components). We describe the developed datasets as well as results showing the effectiveness of the developed datasets.

Given the common ancestry of Dutch and Afrikaans, it is not surprising that they use similar periphrastic constructions to express progressive meaning: aan het (Dutch) and aan die/’t (Afrikaans) lit. ‘at the’; bezig met/(om) te (Dutch) lit. ‘busy with/to’ and besig om te lit. ‘busy to’ (Afrikaans); and so-called cardinal posture verb constructions (zitten/sit ‘sit’, staan ‘stand’, liggen/lê ‘lie’ and lopen/loop ‘walk’), CPV te (‘to’ Dutch) and CPV en (‘and’ Afrikaans). However, these cognate constructions have grammaticalized to different extents. To assess the exact nature of these differences, we analyzed the constructions with respect to overall frequency, collocational range, and transitivity (compatibility with transitive predicates and passivizability). We used two corpora that are equal in size (both about 57 million words) and contain roughly the same types of written text. It turns out that the use of periphrastic progressives is generally more widespread in Afrikaans than in Dutch. As far as grammaticalization is concerned, we found that the Afrikaans aan die- and CPV-constructions, as well as the Dutch bezig- and CPV-constructions, are semantically restricted. In addition, only the Afrikaans besig- and CPV en-constructions allow passivization, which is remarkable for such periphrastic expressions.*