The increasing availability of textual corpora and data fetched from social networks is fuelling a huge production of works based on the model proposed by psychologist Robert Plutchik, often referred simply as the ``Plutchik Wheel''. Related researches range from annotation tasks description to emotions detection tools. Visualisation of such emotions is traditionally carried out using the most popular layouts, as bar plots or tables, which are however sub-optimal. The classic representation of the Plutchik's wheel follows the principles of proximity and opposition between pairs of emotions: spatial proximity in this model is also a semantic proximity, as adjacent emotions elicit a complex emotion (a primary dyad) when triggered together; spatial opposition is a semantic opposition as well, as positive emotions are opposite to negative emotions. The most common layouts fail to preserve both features, not to mention the need of visually allowing comparisons between different corpora in a blink of an eye, that is hard with basic design solutions. We introduce PyPlutchik, a Python library specifically designed for the visualisation of Plutchik's emotions in texts or in corpora. PyPlutchik draws the Plutchik's flower with each emotion petal sized after how much that emotion is detected or annotated in the corpus, also representing three degrees of intensity for each of them. Notably, PyPlutchik allows users to display also primary, secondary, tertiary and opposite dyads in a compact, intuitive way. We substantiate our claim that PyPlutchik outperforms other classic visualisations when displaying Plutchik emotions and we showcase a few examples that display our library's most compelling features.

We present skweak, a versatile, Python-based software toolkit enabling NLP developers to apply weak supervision to a wide range of NLP tasks. Weak supervision is an emerging machine learning paradigm based on a simple idea: instead of labelling data points by hand, we use labelling functions derived from domain knowledge to automatically obtain annotations for a given dataset. The resulting labels are then aggregated with a generative model that estimates the accuracy (and possible confusions) of each labelling function. The skweak toolkit makes it easy to implement a large spectrum of labelling functions (such as heuristics, gazetteers, neural models or linguistic constraints) on text data, apply them on a corpus, and aggregate their results in a fully unsupervised fashion. skweak is especially designed to facilitate the use of weak supervision for NLP tasks such as text classification and sequence labelling. We illustrate the use of skweak for NER and sentiment analysis. skweak is released under an open-source license and is available at: https://github.com/NorskRegnesentral/skweak

Sentiment analysis is an important task in natural language processing (NLP). Most of existing state-of-the-art methods are under the supervised learning paradigm. However, human annotations can be scarce. Thus, we should leverage more weak supervision for sentiment analysis. In this paper, we propose a posterior regularization framework for the variational approach to the weakly supervised sentiment analysis to better control the posterior distribution of the label assignment. The intuition behind the posterior regularization is that if extracted opinion words from two documents are semantically similar, the posterior distributions of two documents should be similar. Our experimental results show that the posterior regularization can improve the original variational approach to the weakly supervised sentiment analysis and the performance is more stable with smaller prediction variance.

The ability to automatically extract Knowledge Graphs (KG) from a given collection of documents is a long-standing problem in Artificial Intelligence. One way to assess this capability is through the task of slot filling. Given an entity query in form of [Entity, Slot, ?], a system is asked to `fill' the slot by generating or extracting the missing value from a relevant passage or passages. This capability is crucial to create systems for automatic knowledge base population, which is becoming in ever-increasing demand, especially in enterprise applications. Recently, there has been a promising direction in evaluating language models in the same way we would evaluate knowledge bases, and the task of slot filling is the most suitable to this intent. The recent advancements in the field try to solve this task in an end-to-end fashion using retrieval-based language models. Models like Retrieval Augmented Generation (RAG) show surprisingly good performance without involving complex information extraction pipelines. However, the results achieved by these models on the two slot filling tasks in the KILT benchmark are still not at the level required by real-world information extraction systems. In this paper, we describe several strategies we adopted to improve the retriever and the generator of RAG in order to make it a better slot filler. Our KGI0 system (available at https://github.com/IBM/retrieve-write-slot-filling) reached the top-1 position on the KILT leaderboard on both T-REx and zsRE dataset with a large margin.

The idea to apply NLP methods to scientific literature seems quite natural. First of all, scientific texts are already well-structured, they contain things like keywords, abstract, as well as well-defined terms. Thus, at the very beginning of COVID pandemic, a research challenge has been launched on Kaggle to analyze scientific papers on the subject. The dataset behind this competition is called CORD (publication), and it contains constantly updated corpus of everything that is published on topics related to COVID. Currently, it contains more than 400000 scientific papers, about half of them - with full text.

Ireland's Data Protection Commission (DPC) is investigating the recent leak of a Facebook user dataset that dates back to 2019. At the start of April, it came out that someone on a hacking forum had made the dataset public, exposing the personal information of about 533 million Facebook users in 106 countries. Depending on the account, there are details about phone numbers, birth dates, email addresses, locations and more. The source of the leak is an oversight Facebook fixed in August 2019. "The DPC, having considered the information provided by Facebook Ireland regarding this matter to date, is of the opinion that one or more provisions of the GDPR and/or the Data Protection Act 2018 may have been, and/or are being, infringed in relation to Facebook Users' personal data," the agency said in a statement spotted by TechCrunch.

Multimodal Sentiment Analysis (MuSe) 2021 is a challenge focusing on the tasks of sentiment and emotion, as well as physiological-emotion and emotion-based stress recognition through more comprehensively integrating the audio-visual, language, and biological signal modalities. The purpose of MuSe 2021 is to bring together communities from different disciplines; mainly, the audio-visual emotion recognition community (signal-based), the sentiment analysis community (symbol-based), and the health informatics community. We present four distinct sub-challenges: MuSe-Wilder and MuSe-Stress which focus on continuous emotion (valence and arousal) prediction; MuSe-Sent, in which participants recognise five classes each for valence and arousal; and MuSe-Physio, in which the novel aspect of `physiological-emotion' is to be predicted. For this years' challenge, we utilise the MuSe-CaR dataset focusing on user-generated reviews and introduce the Ulm-TSST dataset, which displays people in stressful depositions. This paper also provides detail on the state-of-the-art feature sets extracted from these datasets for utilisation by our baseline model, a Long Short-Term Memory-Recurrent Neural Network. For each sub-challenge, a competitive baseline for participants is set; namely, on test, we report a Concordance Correlation Coefficient (CCC) of .4616 CCC for MuSe-Wilder; .4717 CCC for MuSe-Stress, and .4606 CCC for MuSe-Physio. For MuSe-Sent an F1 score of 32.82 % is obtained.

Facebook will not notify the more than half a billion people caught up in a huge leak of personal information, it has said. Over the weekend, it emerged that a vast trove of data on more than 530 million users – containing information including their phone numbers and dates of birth – was being made freely available online. Facebook said that the data was gathered before 2019. It later said that " "malicious actors" had obtained the data prior to September 2019 by "scraping" profiles using a vulnerability in the platform's tool for synching contacts, and that the loophole that allowed them to do so had now been closed. But it said that it did not inform users when the leak happened, and does not have plans to do so now.

Cybersecurity experts revealed a few days ago that over half a billion Facebook users' personal information have been leaked. It's a gold mine of data, which includes users' full names, birthdays, locations and phone numbers. Although Facebook claims that the actual hack happened a couple of years ago, it won't hurt if users made sure their account is not part of the breach and if they are, they should take a few preventive measures to ensure future incidents as messy as this one won't affect them. Australian Security Researcher and HaveIBeenPawned Founder Tony Hunt recently added the 533 million phone numbers exposed in the Facebook data leak to his website. Those worried if their mobile numbers were part of the leak can visit the site and check if their number is there.

The news: The personal data of 533 million Facebook users in more than 106 countries was found to be freely available online last weekend. The data trove, uncovered by security researcher Alon Gal, includes phone numbers, email addresses, home towns, full names, and birth dates. Initially, Facebook claimed that the data leak was previously reported on in 2019 and that it had patched the vulnerability that caused it that August. But in fact, it appears that Facebook did not properly disclose the breach at the time. It only finally acknowledged it on Tuesday April 6 in a blog post by product management director Mike Clark.