More than a month has passed since our incredible first-time experience attending the 2022 World Economic Forum Annual Meeting in Davos. The trip to Davos, the people I met, the conversations I had, and the events I attended helped me understand that Citibeats is not a fancy personal project about how to use technology for social good. To me, more than a company (and a great one, thanks to all my colleagues), Citibeats is an idea. And this idea, now more than ever is absolutely necessary. I think there has never been a historical period so lucidly conscious of the change.

Acumen Research and Consulting recently published report titled "Artificial Intelligence in Manufacturing Market Size, Share, Analysis Report and Region Forecast, 2022 - 2030" BEIJING, July 11, 2022 (GLOBE NEWSWIRE) -- The Global Artificial Intelligence in Manufacturing Market size accounted for USD 2,963 Million in 2021 and is estimated to reach USD 78,744 Million by 2030. The rising volume of complex data sets is the leading factor boosting the global artificial intelligence in manufacturing market revenue. Our worldwide artificial intelligence in manufacturing industry analysis suggests that the manufacturers require artificial intelligence (AI) in their facilities due to the surging need for enhanced productivity and automation. AI is being used by manufacturers to enhance day-to-day operations, introduce new products, personalize designs, and forecast future financials. According to an MIT survey, about 60% of industry players are already using artificial intelligence.

Kaizen Gaming is the leading GameTech company in Greece and one of the fastest-growing in Europe, with the Stoiximan brand in Greece and Cyprus and Betano in Germany, Romania, Bulgaria, Portugal, Brazil, Chile and Peru. Our aim is to leverage cutting-edge Technology in order to provide the optimum experience to those who trust us for their entertainment. At Kaizen, our aim is to make data driven decisions in order to automate our services while also focusing on offering tailored customer experiences. Our machine learning team is dedicated to this mission by building a variety of models, from binary classification tasks up to recommendation systems. We focus on transforming business needs into production applications and we cover a wide range of business sectors utilizing different data types and handling a broad project diversity.

We introduce ParaNames, a multilingual parallel name resource consisting of 118 million names spanning across 400 languages. Names are provided for 13.6 million entities which are mapped to standardized entity types (PER/LOC/ORG). Using Wikidata as a source, we create the largest resource of this type to-date. We describe our approach to filtering and standardizing the data to provide the best quality possible. ParaNames is useful for multilingual language processing, both in defining tasks for name translation/transliteration and as supplementary data for tasks such as named entity recognition and linking. We demonstrate an application of ParaNames by training a multilingual model for canonical name translation to and from English. Our resource is released under a Creative Commons license (CC BY 4.0) at https://github.com/bltlab/paranames.

We investigate whether it is feasible to remove gendered information from resumes to mitigate potential bias in algorithmic resume screening. Using a corpus of 709k resumes from IT firms, we first train a series of models to classify the self-reported gender of the applicant, thereby measuring the extent and nature of gendered information encoded in resumes. We then conduct a series of gender obfuscation experiments, where we iteratively remove gendered information from resumes. Finally, we train a resume screening algorithm and investigate the trade-off between gender obfuscation and screening algorithm performance. Results show: (1) There is a significant amount of gendered information in resumes. (2) Lexicon-based gender obfuscation method (i.e. removing tokens that are predictive of gender) can reduce the amount of gendered information to a large extent. However, after a certain point, the performance of the resume screening algorithm starts suffering. (3) General-purpose gender debiasing methods for NLP models such as removing gender subspace from embeddings are not effective in obfuscating gender.

As AI systems become increasingly powerful and pervasive, there are growing concerns about machines' morality or a lack thereof. Yet, teaching morality to machines is a formidable task, as morality remains among the most intensely debated questions in humanity, let alone for AI. Existing AI systems deployed to millions of users, however, are already making decisions loaded with moral implications, which poses a seemingly impossible challenge: teaching machines moral sense, while humanity continues to grapple with it. To explore this challenge, we introduce Delphi, an experimental framework based on deep neural networks trained directly to reason about descriptive ethical judgments, e.g., "helping a friend" is generally good, while "helping a friend spread fake news" is not. Empirical results shed novel insights on the promises and limits of machine ethics; Delphi demonstrates strong generalization capabilities in the face of novel ethical situations, while off-the-shelf neural network models exhibit markedly poor judgment including unjust biases, confirming the need for explicitly teaching machines moral sense. Yet, Delphi is not perfect, exhibiting susceptibility to pervasive biases and inconsistencies. Despite that, we demonstrate positive use cases of imperfect Delphi, including using it as a component model within other imperfect AI systems. Importantly, we interpret the operationalization of Delphi in light of prominent ethical theories, which leads us to important future research questions.

More and more investors and machine learning models rely on social media (e.g., Twitter and Reddit) to gather real-time information and sentiment to predict stock price movements. Although text-based models are known to be vulnerable to adversarial attacks, whether stock prediction models have similar vulnerability is underexplored. In this paper, we experiment with a variety of adversarial attack configurations to fool three stock prediction victim models. We address the task of adversarial generation by solving combinatorial optimization problems with semantics and budget constraints. Our results show that the proposed attack method can achieve consistent success rates and cause significant monetary loss in trading simulation by simply concatenating a Figure 1: An example of word-replacement adversarial perturbed but semantically similar tweet.

The graph represents a network of 2,666 Twitter users whose tweets in the requested range contained "datamining", or who were replied to or mentioned in those tweets. The network was obtained from the NodeXL Graph Server on Monday, 11 July 2022 at 06:54 UTC. The requested start date was Monday, 11 July 2022 at 00:01 UTC and the maximum number of days (going backward) was 14. The maximum number of tweets collected was 7,500. The tweets in the network were tweeted over the 13-day, 22-hour, 59-minute period from Monday, 27 June 2022 at 01:01 UTC to Monday, 11 July 2022 at 00:00 UTC.

Deep learning-based approaches for content-based image retrieval (CBIR) of CT liver images is an active field of research, but suffers from some critical limitations. First, they are heavily reliant on labeled data, which can be challenging and costly to acquire. Second, they lack transparency and explainability, which limits the trustworthiness of deep CBIR systems. We address these limitations by (1) proposing a self-supervised learning framework that incorporates domain-knowledge into the training procedure and (2) providing the first representation learning explainability analysis in the context of CBIR of CT liver images. Results demonstrate improved performance compared to the standard self-supervised approach across several metrics, as well as improved generalisation across datasets. Further, we conduct the first representation learning explainability analysis in the context of CBIR, which reveals new insights into the feature extraction process. Lastly, we perform a case study with cross-examination CBIR that demonstrates the usability of our proposed framework. We believe that our proposed framework could play a vital role in creating trustworthy deep CBIR systems that can successfully take advantage of unlabeled data.

Current pre-trained language models (PLM) are typically trained with static data, ignoring that in real-world scenarios, streaming data of various sources may continuously grow. This requires PLMs to integrate the information from all the sources in a lifelong manner. Although this goal could be achieved by exhaustive pre-training on all the existing data, such a process is known to be computationally expensive. To this end, we propose ELLE, aiming at efficient lifelong pre-training for emerging data. Specifically, ELLE consists of (1) function preserved model expansion, which flexibly expands an existing PLM's width and depth to improve the efficiency of knowledge acquisition; and (2) pre-trained domain prompts, which disentangle the versatile knowledge learned during pre-training and stimulate the proper knowledge for downstream tasks. We experiment ELLE with streaming data from 5 domains on BERT and GPT. The results show the superiority of ELLE over various lifelong learning baselines in both pre-training efficiency and downstream performances. The codes are publicly available at https://github.com/thunlp/ELLE.