On Friday, Elon Musk announced he was pausing his $45bn purchase of Twitter because he had only just discovered some of the accounts on the site were fake. But that's not the strangest thing that has happened to the beleaguered social media platform this week. Because on Tuesday the current top brass, perhaps trying to demonstrate their vision for the site, released a Super Nintendo-style browser game that recaps Twitter's private policy. The platform unveiled Twitter Data Dash, which plays like a vintage side-scrolling platformer that's been draped with a healthy dose of disinformation anxiety. You take control of a blue-hued puppy named Data and are tasked with retrieving five bones hidden in each of the game's day-glo urban environments.

Call Center sentiment analysis is the processing of data by identifying the natural nuance of customer context and analyzing data to make customer service more empathetic. If you are employed in Call Center, the following scenario might be familiar: You get a call from a client and hear their words with stress. The cause for such a cataclysmic reaction: They got a bad rating for their products or business. Some of those reviews might be negative, formal, and neutral. Knowing what someone meant can be tricky unless you understand their emotional quotient.

Digital rights non-profit Fight for the Future and 27 human rights organizations have written an open letter to Zoom, asking the company not to continue exploring the use of AI that can analyze emotions in its video conferencing platform. The groups wrote the letter in response to a Protocol report that said Zoom is actively researching how to incorporate emotion AI into its product in the future. It's part of a larger piece examining how companies have started using artificial intelligence to detect the emotional state of a potential client during sales calls. The pandemic made video conferences a lot more common around the world. Sales people have been finding it hard to gauge how receptive potential clients are to their products and services, though, without the capability to read their body language through the screen.

Euphemisms have not received much attention in natural language processing, despite being an important element of polite and figurative language. Euphemisms prove to be a difficult topic, not only because they are subject to language change, but also because humans may not agree on what is a euphemism and what is not. Nevertheless, the first step to tackling the issue is to collect and analyze examples of euphemisms. We present a corpus of potentially euphemistic terms (PETs) along with example texts from the GloWbE corpus. Additionally, we present a subcorpus of texts where these PETs are not being used euphemistically, which may be useful for future applications. We also discuss the results of multiple analyses run on the corpus. Firstly, we find that sentiment analysis on the euphemistic texts supports that PETs generally decrease negative and offensive sentiment. Secondly, we observe cases of disagreement in an annotation task, where humans are asked to label PETs as euphemistic or not in a subset of our corpus text examples. We attribute the disagreement to a variety of potential reasons, including if the PET was a commonly accepted term (CAT).

The Covid-19 pandemic has led to infodemic of low quality information leading to poor health decisions. Combating the outcomes of this infodemic is not only a question of identifying false claims, but also reasoning about the decisions individuals make. In this work we propose a holistic analysis framework connecting stance and reason analysis, and fine-grained entity level moral sentiment analysis. We study how to model the dependencies between the different level of analysis and incorporate human insights into the learning process. Experiments show that our framework provides reliable predictions even in the low-supervision settings.

Transformers are one of the most exciting concepts in Natural Language Processing. This article is a guide on working with pre-trained models that use transformers. A transformer model is a neural…

Aspect-level sentiment analysis aims to determine the sentiment polarity towards a specific target in a sentence. The main challenge of this task is to effectively model the relation between targets and sentiments so as to filter out noisy opinion words from irrelevant targets. Most recent efforts capture relations through target-sentiment pairs or opinion spans from a word-level or phrase-level perspective. Based on the observation that targets and sentiments essentially establish relations following the grammatical hierarchy of phrase-clause-sentence structure, it is hopeful to exploit comprehensive syntactic information for better guiding the learning process. Therefore, we introduce the concept of Scope, which outlines a structural text region related to a specific target. To jointly learn structural Scope and predict the sentiment polarity, we propose a hybrid graph convolutional network (HGCN) to synthesize information from constituency tree and dependency tree, exploring the potential of linking two syntax parsing methods to enrich the representation. Experimental results on four public datasets illustrate that our HGCN model outperforms current state-of-the-art baselines.

By now, most of Twitter's 217 million daily active users have probably heard the news: Elon Musk -- the world's richest person, CEO of Tesla and SpaceX and a prolific Internet poster -- has reached an agreement to buy the social network for about $44 billion.

Originally published on Towards AI the World's Leading AI and Technology News and Media Company. If you are building an AI-related product or service, we invite you to consider becoming an AI sponsor. At Towards AI, we help scale AI and technology startups. Let us help you unleash your technology to the masses. "Your most unhappy customers are your greatest source of learning."

Researchers from the UPV/EHU-University of the Basque Country, the UNED (National Distance Education University) and Elhuyar have created the VIGICOVID system, thanks to Supera COVID-19 (Overcoming COVID-19) funding by the CRUE (Association of Spanish Universities). This system addresses the need to search for answers in the avalanche of information generated by all the research conducted across the world relating to the pandemic. By means of artificial intelligence, the system displays the answers found in a set of scientific articles in an orderly fashion, and uses natural language questions and answers. The global bio-health research community is making a tremendous effort to generate knowledge relating to COVID-19 and SARS-CoV-2. In practice, this effort means a huge, very rapid production of scientific publications, which makes it difficult to consult and analyze all the information. That is why experts and decision-making bodies need to be provided with information systems to enable them to acquire the knowledge they need.