The experimental landscape in natural language processing for social media is too fragmented. Each year, new shared tasks and datasets are proposed, ranging from classics like sentiment analysis to irony detection or emoji prediction. Therefore, it is unclear what the current state of the art is, as there is no standardized evaluation protocol, neither a strong set of baselines trained on such domainspecific data. In this paper, we propose a new evaluation framework (TWEETEVAL) consisting of seven heterogeneous Twitter-specific classification tasks. We also provide a strong set of baselines as starting point, and compare different language modeling pre-training strategies. Our initial experiments show the effectiveness of starting off with existing pretrained generic language models, and continue training them on Twitter corpora.

Everyday billions of multimodal posts containing both images and text are shared in social media sites such as Snapchat, Twitter or Instagram. This combination of image and text in a single message allows for more creative and expressive forms of communication, and has become increasingly common in such sites. This new paradigm brings new challenges for natural language understanding, as the textual component tends to be shorter, more informal, and often is only understood if combined with the visual context. In this paper, we explore the task of name tagging in multimodal social media posts. We start by creating two new multimodal datasets: one based on Twitter posts 1 and the other based on Snapchat captions (exclusively submitted to public and crowdsourced stories). We then propose a novel model based on Visual Attention that not only provides deeper visual understanding on the decisions of the model, but also significantly outperforms other state-of-theart baseline methods for this task. 2 * * This work was mostly done during the first author's internship at Snap Research. 1 The Twitter data and associated images presented in this paper were downloaded from https://archive.org/ details/twitterstream 2 We will make the annotations on Twitter data available for research purpose upon request.

We introduce a new task called Multimodal Named Entity Recognition (MNER) for noisy user-generated data such as tweets or Snapchat captions, which comprise short text with accompanying images. These social media posts often come in inconsistent or incomplete syntax and lexical notations with very limited surrounding textual contexts, bringing significant challenges for NER. To this end, we create a new dataset for MNER called SnapCaptions (Snapchat image-caption pairs submitted to public and crowd-sourced stories with fully annotated named entities). We then build upon the state-of-the-art Bi-LSTM word/character based NER models with 1) a deep image network which incorporates relevant visual context to augment textual information, and 2) a generic modality-attention module which learns to attenuate irrelevant modalities while amplifying the most informative ones to extract contexts from, adaptive to each sample and token. The proposed MNER model with modality attention significantly outperforms the state-of-the-art text-only NER models by successfully leveraging provided visual contexts, opening up potential applications of MNER on myriads of social media platforms.

We introduce the new Multimodal Named Entity Disambiguation (MNED) task for multimodal social media posts such as Snapchat or Instagram captions, which are composed of short captions with accompanying images. Social media posts bring significant challenges for disambiguation tasks because 1) ambiguity not only comes from polysemous entities, but also from inconsistent or incomplete notations, 2) very limited context is provided with surrounding words, and 3) there are many emerging entities often unseen during training. To this end, we build a new dataset called SnapCaptionsKB, a collection of Snapchat image captions submitted to public and crowd-sourced stories, with named entity mentions fully annotated and linked to entities in an external knowledge base. We then build a deep zeroshot multimodal network for MNED that 1) extracts contexts from both text and image, and 2) predicts correct entity in the knowledge graph embeddings space, allowing for zeroshot disambiguation of entities unseen in training set as well. The proposed model significantly outperforms the stateof-the-art text-only NED models, showing efficacy and potentials of the MNED task.

Parkinson's disease (PD) presents several motor signs, including tremor and bradykinesia. However, these signs can also be found in other motor disorders and in neurologically healthy older adults. The incidence of bradykinesia in PD is relatively high in all stages of the disorder, even when compared to tremor. Thus, this research proposes an objective assessment of bradykinesia in patients with PD (G : 15 older adults with Parkinson's disease, 65.3 ± 9.1 years) and older adults (G: 12 healthy older adults, 60.1 ± 6.1 years). The severity of bradykinesia in the participants of G was assessed using the Unified Parkinson's Disease Rating Scale. Movement and muscular activity were detected by means of inertial (accelerometer, gyroscope, magnetometer) and electromyographic sensors while the participants performed wrist extension against gravity with the forearm on pronation. Mean and standard error of inertial and electromyographic signal parameters could discriminate PD patients from healthy older adults (p value<0.05). In discriminating patients with PD from healthy older adults, the mean sensitivity and specificity were respectively 86.67 and 83.33%. The discrimination between the groups, based on the objective evaluation of bradykinesia, may contribute to the accurate diagnosis of PD and to the monitoring of therapies to control parkinsonian bradykinesia, and opens the possibility for further comparative studies considering individuals suffering from other motor disorders.