Inflectional variation is a common feature of World Englishes such as Colloquial Singapore English and African American Vernacular English. Although comprehension by human readers is usually unimpaired by non-standard inflections, current NLP systems are not yet robust. We propose Base-Inflection Encoding (BITE), a method to tokenize English text by reducing inflected words to their base forms before reinjecting the grammatical information as special symbols. Fine-tuning pretrained NLP models for downstream tasks using our encoding defends against inflectional adversaries while maintaining performance on clean data. Models using BITE generalize better to dialects with non-standard inflections without explicit training and translation models converge faster when trained with BITE. Finally, we show that our encoding improves the vocabulary efficiency of popular data-driven subword tokenizers. Since there has been no prior work on quantitatively evaluating vocabulary efficiency, we propose metrics to do so.

Clinical machine learning is increasingly multimodal, collected in both structured tabular formats and unstructured forms such as freetext. We propose a novel task of exploring fairness on a multimodal clinical dataset, adopting equalized odds for the downstream medical prediction tasks. To this end, we investigate a modality-agnostic fairness algorithm - equalized odds post processing - and compare it to a text-specific fairness algorithm: debiased clinical word embeddings. Despite the fact that debiased word embeddings do not explicitly address equalized odds of protected groups, we show that a text-specific approach to fairness may simultaneously achieve a good balance of performance and classical notions of fairness. We hope that our paper inspires future contributions at the critical intersection of clinical NLP and fairness. The full source code is available here: https://github.com/johntiger1/multimodal_fairness

In this paper we show that simple semidefinite programs inspired by degree $4$ SOS can exactly solve the tensor nuclear norm, tensor decomposition, and tensor completion problems on tensors with random asymmetric components. More precisely, for tensor nuclear norm and tensor decomposition, we show that w.h.p. these semidefinite programs can exactly find the nuclear norm and components of an $(n\times n\times n)$-tensor $\mathcal{T}$ with $m\leq n^{3/2}/polylog(n)$ random asymmetric components. For tensor completion, we show that w.h.p. the semidefinite program introduced by Potechin \& Steurer (2017) can exactly recover an $(n\times n\times n)$-tensor $\mathcal{T}$ with $m$ random asymmetric components from only $n^{3/2}m\, polylog(n)$ randomly observed entries. This gives the first theoretical guarantees for exact tensor completion in the overcomplete regime. This matches the best known results for approximate versions of these problems given by Barak \& Moitra (2015) for tensor completion, and Ma, Shi \& Steurer (2016) for tensor decomposition.

Artificial intelligence is presenting considerable advancements in the generation of convincing and fascinating images of unquestionably realistic features that are almost impossible to be classified as belonging to people who do not exist anywhere in the world. If we compare the faces produced no more than five years ago and those published recently, the improvements are incredible. With the new GANs -- Generative Adversarial Networks, algorithmic architectures that use two neural networks, competing one against the other (thus the term "adversarial") to generate new, synthetic instances of data, these synthetic faces are easily customizable and editable, making them so credible thanks to particular effects. Applying a mix of features from real images such as the shades of the skin and the color of the hair, for example, to the fake ones is possible to generate a virtual population that does not exist in the real world. But why the use of images of non-existent people, created by an algorithm, instead of photos of real people can be so exciting?

CallMiner, the leading provider of speech and customer interaction analytics, today announced that African Bank has selected the CallMiner Eureka platform to better extract and operationalise insights from voice-based customer interactions. "With CallMiner, we will be able to create an even stronger customer experience by analyzing what our customers need, how they feel, and what we need to do, to deliver exceptional value and service," "African Bank is fast becoming a pioneering data-driven South African organization and our partnership with CallMiner is another part of our journey to make our organisation synonymous with data excellence and to enhance our digital transformation. At the core of our data strategy are our customers – to create an inclusive, personalised banking experience for them and we believe CallMiner provides another tool to bring us one step closer to achieving this." CallMiner Eureka leverages artificial intelligence and machine learning to analyse every customer interaction, across all channels, to automatically uncover actionable insights and intelligence that organisations can use to improve business decisions. African Bank will start analysing voice interactions through its three contact centres initially.

Researchers counted over 1.8 billion trees and shrubs in the 1.3 million square kilometer (501,933 square miles) area that covers the western-most portion of the Sahara Desert, the Sahel, and what are known as sub-humid zones of West Africa. "We were very surprised to see that quite a few trees actually grow in the Sahara Desert, because up until now, most people thought that virtually none existed," says Martin Brandt, professor in the geosciences and natural resource management department at the University of Copenhagen and lead author of the study in Nature. "We counted hundreds of millions of trees in the desert alone. Doing so wouldn't have been possible without this technology. Indeed, I think it marks the beginning of a new scientific era."

But the president's aspirations have long run into resistance, as his own national security officials argued that abandonment of such troubled countries could have catastrophic consequences -- such as when the United States pulled out of Iraq at the end of 2011, leaving a vacuum that fostered the rise of the Islamic State in Iraq and Syria. Mr. Trump has also repeatedly pushed to withdraw from Syria, but several hundred U.S. troops remain stationed there, partly to protect coveted oil fields held by American-backed Syrian Kurdish allies from being seized by the government of President Bashar al-Assad of Syria. The current deliberations over withdrawals would not affect those in Syria, officials said. The plan under discussion to pull out of Somalia is said to not apply to U.S. forces stationed in nearby Kenya and Djibouti, where American drones that carry out airstrikes in Somalia are based, according to officials familiar with the internal deliberations who spoke on the condition of anonymity. Keeping those air bases would mean retaining the military's ability to use drones to attack militants with the Shabab, the Qaeda-linked terrorist group -- at least those deemed to pose a threat to American interests.

Amazon Web Services' new visual data preparation tool for AWS Glue allows users to clean and normalize data with an interactive point-and-click visual interface without writing custom code. AWS Glue DataBrew helps data scientists and data analysts get the data ready for analytics and machine learning (ML) 80 percent quicker than traditional data preparation approaches, according to the cloud provider, which made the tool generally available on Wednesday. The new offering builds on AWS Glue, which AWS generally released in April of 2017. AWS Glue is a serverless, fully managed, extract, transform and load (ETL) service to categorize, clean, enrich and move data between various data stores. It has a central data repository called the AWS Glue Data Catalog, an ETL engine that generates Python code automatically and a flexible scheduler to handle dependency resolution, job monitoring and retries.

Knowledge graphs have been demonstrated to be an effective tool for numerous intelligent applications. However, a large amount of valuable knowledge still exists implicitly in the knowledge graphs. To enrich the existing knowledge graphs, recent years witness that many algorithms for link prediction and knowledge graphs embedding have been designed to infer new facts. But most of these studies focus on the static knowledge graphs and ignore the temporal information that reflects the validity of knowledge. Developing the model for temporal knowledge graphs completion is an increasingly important task. In this paper, we build a new tensor decomposition model for temporal knowledge graphs completion inspired by the Tucker decomposition of order 4 tensor. We demonstrate that the proposed model is fully expressive and report state-of-the-art results for several public benchmarks. Additionally, we present several regularization schemes to improve the strategy and study their impact on the proposed model. Experimental studies on three temporal datasets (i.e. ICEWS2014, ICEWS2005-15, GDELT) justify our design and demonstrate that our model outperforms baselines with an explicit margin on link prediction task.

When creating or rather crafting a chatbot conversation we as designers must draw inspiration and guidance from real-world conversations. Elements of human conversation should be identified and abstracted to be incorporated in our chatbot conversation. General rules and concepts of human conversations must be derived and implemented via technically astute means. Below I list 10 elements of human conversation which can be incorporated in a Conversational AI interface. Conversational designers want users to speak to their chatbot as to a human…hence it is time for the chatbot to converse more human like. Christoph Niemann has fascinating ideas on abstraction and when visual design becomes too abstract. The speaker introduces a topic, subsequently the speaker introduces a second topic, another story that seems to be unrelated.