Community clinics say the easing of restrictions on telehealth during the pandemic has made it possible for health workers to connect with hard-to-reach patients via a phone call -- people who are poor, elderly or live in remote areas, and don't have access to a computer or cell phone with video capability. Community clinics say the easing of restrictions on telehealth during the pandemic has made it possible for health workers to connect with hard-to-reach patients via a phone call -- people who are poor, elderly or live in remote areas, and don't have access to a computer or cell phone with video capability. Caswell County, where William Crumpton works, runs along the northern edge of North Carolina and is a rural landscape of mostly former tobacco farms and the occasional fast-food restaurant. "There are wide areas where cell phone signals are just nonexistent," Crumpton says. "Things like satellite radio are even a challenge."

Although reading books and watching lectures is a great way to learn analytics – it is best to start doing. However, it can be quite tricky to start doing when it comes to languages such as Python and R if someone does not have a coding background. Not only do you need to know what you are doing in terms of analytical procedures, but you also need to understand the nuances of programming languages which adds onto the list of things to learn to just get started. Therefore, the best middle ground between knowledge acquisition (books, videos, etc.) and conducting advanced analytics (Python, R, etc.) is by using open-source analytics software. These types of software are great for both knowledge acquisition and actually doing analysis as documentation is built into the software and you can start doing relatively complex tasks with only mouse clicks.

Knowledge enriched language representation learning has shown promising performance across various knowledge-intensive NLP tasks. However, existing knowledge based language models are all trained with monolingual knowledge graph data, which limits their application to more languages. In this work, we present a novel framework to pretrain knowledge based multilingual language models (KMLMs). We first generate a large amount of code-switched synthetic sentences and reasoning-based multilingual training data using the Wikidata knowledge graphs. Then based on the intra-and inter-sentence structures of the generated data, we design pretraining tasks to facilitate knowledge learning, which allows the language models to not only memorize the factual knowledge but also learn useful logical patterns. Our pretrained KMLMs demonstrate significant performance improvements on a wide range of knowledge-intensive cross-lingual NLP tasks, including named entity recognition, factual knowledge retrieval, relation classification, and a new task designed by us, namely, logic reasoning. Our code and pretrained language models will be made publicly available. Pretrained language models (PTLMs) such as BERT (Devlin et al., 2019) and RoBERTa (Liu et al., 2019) have achieved superior performances on a wide range of natural language processing (NLP) tasks.

The North Carolina Agriculture and Technical State University (NC A&T) in collaboration with Georgia Tech Research Institute (GTRI) has developed methodologies for creating simulation-based technology tools that are capable of inferring the perceptions and behavioral states of autonomous systems. These methodologies have the potential to provide the Test and Evaluation (T&E) community at the Department of Defense (DoD) with a greater insight into the internal processes of these systems. The methodologies use only external observations and do not require complete knowledge of the internal processing of and/or any modifications to the system under test. This paper presents an example of one such simulation-based technology tool, named as the Data-Driven Intelligent Prediction Tool (DIPT). DIPT was developed for testing a multi-platform Unmanned Aerial Vehicle (UAV) system capable of conducting collaborative search missions. DIPT's Graphical User Interface (GUI) enables the testers to view the aircraft's current operating state, predicts its current target-detection status, and provides reasoning for exhibiting a particular behavior along with an explanation of assigning a particular task to it.

Cybersecurity has been a concern for quite a while now. In the latest years, cyberattacks have been increasing in size and complexity, fueled by significant advances in technology. Nowadays, there is an unavoidable necessity of protecting systems and data crucial for business continuity. Hence, many intrusion detection systems have been created in an attempt to mitigate these threats and contribute to a timelier detection. This work proposes an interpretable and explainable hybrid intrusion detection system, which makes use of artificial intelligence methods to achieve better and more long-lasting security. The system combines experts' written rules and dynamic knowledge continuously generated by a decision tree algorithm as new shreds of evidence emerge from network activity.

While talking about Artificial Intelligence (AI) in healthcare might sound futuristic, the first proof of concept for AI application took place in the late 1950s1. In the 1970s, researchers at Stanford developed the MYCIN program to help doctors identify blood infections.2 At Intel, we've had the opportunity to see many different types of AI applications in use by our partners in the healthcare industry, from AI-enabled robots that can help clean hospital rooms to algorithms that can perform real-time inference on endoscopic cameras. Many of these AI implementations rely on edge computing, or the ability to process and compute data close to where it originates -- either on a network-connected device or right next to the device. AI at the edge means that data can be processed and analyzed quickly -- before it goes to the cloud or a server for storage.

Most existing approaches for Knowledge Base Question Answering (KBQA) focus on a specific underlying knowledge base either because of inherent assumptions in the approach, or because evaluating it on a different knowledge base requires non-trivial changes. However, many popular knowledge bases share similarities in their underlying schemas that can be leveraged to facilitate generalization across knowledge bases. To achieve this generalization, we introduce a KBQA framework based on a 2-stage architecture that explicitly separates semantic parsing from the knowledge base interaction, facilitating transfer learning across datasets and knowledge graphs. We show that pretraining on datasets with a different underlying knowledge base can nevertheless provide significant performance gains and reduce sample complexity. Our approach achieves comparable or state-of-the-art performance for LC-QuAD (DBpedia), WebQSP (Freebase), SimpleQuestions (Wikidata) and MetaQA (Wikimovies-KG).

This paper considers the learning of Boolean rules in either disjunctive normal form (DNF, OR-of-ANDs, equivalent to decision rule sets) or conjunctive normal form (CNF, AND-of-ORs) as an interpretable model for classification. An integer program is formulated to optimally trade classification accuracy for rule simplicity. We also consider the fairness setting and extend the formulation to include explicit constraints on two different measures of classification parity: equality of opportunity and equalized odds. Column generation (CG) is used to efficiently search over an exponential number of candidate clauses (conjunctions or disjunctions) without the need for heuristic rule mining. This approach also bounds the gap between the selected rule set and the best possible rule set on the training data. To handle large datasets, we propose an approximate CG algorithm using randomization. Compared to three recently proposed alternatives, the CG algorithm dominates the accuracy-simplicity trade-off in 8 out of 16 datasets. When maximized for accuracy, CG is competitive with rule learners designed for this purpose, sometimes finding significantly simpler solutions that are no less accurate. Compared to other fair and interpretable classifiers, our method is able to find rule sets that meet stricter notions of fairness with a modest trade-off in accuracy.

Patent data have been utilized for engineering design research for long because it contains massive amount of design information. Recent advances in artificial intelligence and data science present unprecedented opportunities to mine, analyse and make sense of patent data to develop design theory and methodology. Herein, we survey the patent-for-design literature by their contributions to design theories, methods, tools, and strategies, as well as different forms of patent data and various methods. Our review sheds light on promising future research directions for the field.

Possibilistic logic is the most popular approach to represent and reason with uncertain and partially inconsistent knowledge. Regarding normal forms, the encoding of real-world problems does usually not result in a clausal formula and although a possibility nonclausal formula is theoretically equivalent to some possibilistic clausal formula [26, 22], approaches needing clausal form transformations are practically infeasible or have experimentally shown to be highly inefficient as discussed below. Two kinds of clausal form transformation are known: (1) one is based on the repetitive application of the distributive laws to the input non-clausal formula until a logically equivalent clausal formula is obtained; and (2) the other transformation, Tsetin-transformation [59], is based on recursively substituting sub-formulas in the input non-clausal formula by fresh literals until obtaining an equi-satisfiable, but not equivalent, clausal formula.