We present ElPiGraph, a method for approximating data distributions having non-trivial topological features such as the existence of excluded regions or branching structures. Unlike many existing methods, ElPiGraph is not based on the construction of a k-nearest neighbour graph, a procedure that can perform poorly in the case of multidimensional and noisy data. Instead, ElPiGraph constructs elastic principal graphs in a more robust way by minimizing elastic energy, applying graph grammars and explicitly controlling topological complexity. Using trimmed approximation error function makes ElPiGraph extremely robust to the presence of background noise without decreasing computational performance and allows it to deal with complex cases of manifold learning (for example, ElPiGraph can learn disconnected intersecting manifolds). Thanks to the quasi-quadratic nature of the elastic function, ElPiGraph performs almost as fast as a simple k-means clustering and, therefore, is much more scalable than alternative methods, and can work on large datasets containing millions of high dimensional points on a personal computer. The excellent performance of the method opens the possibility to apply resampling and to approximate complex data structures via principal graph ensembles which can be used to construct consensus principal graphs. ElPiGraph is currently implemented in five programming languages and accompanied by a graphical user interface, which makes it a versatile tool to deal with complex data in various fields from molecular biology, where it can be used to infer pseudo-time trajectories from single-cell RNASeq, to astronomy, where it can be used to approximate complex structures in the distribution of galaxies.

The question whether an ontology can safely be replaced by another, possibly simpler, one is fundamental for many ontology engineering and maintenance tasks. It underpins, for example, ontology versioning, ontology modularization, forgetting, and knowledge exchange. What safe replacement means depends on the intended application of the ontology. If, for example, it is used to query data, then the answers to any relevant ontology-mediated query should be the same over any relevant data set; if, in contrast, the ontology is used for conceptual reasoning, then the entailed subsumptions between concept expressions should coincide. This gives rise to different notions of ontology inseparability such as query inseparability and concept inseparability, which generalize corresponding notions of conservative extensions. We survey results on various notions of inseparability in the context of description logic ontologies, discussing their applications, useful model-theoretic characterizations, algorithms for determining whether two ontologies are inseparable (and, sometimes, for computing the difference between them if they are not), and the computational complexity of this problem.

Abstract--This paper presents the statistical model for Ticker [1], a novel probabilistic stereophonic single-switch text entry method for visually-impaired users with motor disabilities who rely on single-switch scanning systems to communicate. All terminology and notation are defined in [1]. In Figure 1(a) a typical composite audio sequence that can be presented to the user is shown, where the composite sequence consists of two repetitions of the alphabet. In Ticker, the user selects one letter at a time when listening to such a sequence. In the shown example, the user can click twice per letter. The second repetition occurs in a different order than the first, which allows one to infer the intentional letter selection more accurately. The system does not explicitly make any selection after a click is received; instead the system accumulates evidence. After one or more clicks are received, the system internally updates the posterior word probabilities. It will then proceed to play the composite sequence again for the next letter. When the posterior probability of any word in a predefined dictionary is above a certain threshold, that word is selected.

The cognitive framework of conceptual spaces bridges the gap between symbolic and subsymbolic AI by proposing an intermediate conceptual layer where knowledge is represented geometrically. There are two main approaches for obtaining the dimensions of this conceptual similarity space: using similarity ratings from psychological experiments and using machine learning techniques. In this paper, we propose a combination of both approaches by using psychologically derived similarity ratings to constrain the machine learning process. This way, a mapping from stimuli to conceptual spaces can be learned that is both supported by psychological data and allows generalization to unseen stimuli. The results of a first feasibility study support our proposed approach.

Dialogue policy transfer enables us to build dialogue policies in a target domain with little data by leveraging knowledge from a source domain with plenty of data. Dialogue sentences are usually represented by speech-acts and domain slots, and the dialogue policy transfer is usually achieved by assigning a slot mapping matrix based on human heuristics. However, existing dialogue policy transfer methods cannot transfer across dialogue domains with different speech-acts, for example, between systems built by different companies. Also, they depend on either common slots or slot entropy, which are not available when the source and target slots are totally disjoint and no database is available to calculate the slot entropy. To solve this problem, we propose a Policy tRansfer across dOMaIns and SpEech-acts (PROMISE) model, which is able to transfer dialogue policies across domains with different speech-acts and disjoint slots. The PROMISE model can learn to align different speech-acts and slots simultaneously, and it does not require common slots or the calculation of the slot entropy. Experiments on both real-world dialogue data and simulations demonstrate that PROMISE model can effectively transfer dialogue policies across domains with different speech-acts and disjoint slots.

Google Maps is a triumph of artificial intelligence in action, with the ability to guide us from one place to another using some impressive machine learning technology. But while the routing part of Google Maps doesn't need too many humans in the mix, manually tracing the roads on the aerial images to make them machine usable is incredibly time-consuming and mundane. As a result, even with thousands of hours spent on this task, Google employees still haven't managed to map the majority of the 20 million-plus miles of roadways that stretch around the world. Fortunately, researchers from the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (CSAIL) may have come up with a solution. They developed an automated method to build roadmaps which is 45 percent more accurate than existing methods.

Many of us now use small doses of AI in everyday life (like Siri, Google Assistant, Alexa, and everything smart home), but hotels are putting this once-sci-fi technology to more widespread use. From concierge robots to personalized rooms to lively chatbots, your next holiday may include help from some artificially intellectualized friends. While you may miss, say, the smile or handshake you get from their human counterparts, these systems can create hyper-personalized experiences and comprehensively upgrade the level of service during your stay. Keep an eye out for these features at your next check-in. Some hotels are using robots to beef up customer service.

The 526th Intelligence Squadron at Nellis Air Force Base, Nevada, hosted an Artificial Intelligence and Design Thinking seminar at AFWERX Vegas. Ian, superintendent of the 9th Intelligence Squadron at Beale AFB, California, introduced more than 100 Airmen, contractors and Department of Defense employees to the fundamentals of design thinking, artificial intelligence and cutting-edge computer technology. "I want to expose you to the way we do business in (intelligence, surveillance and reconnaissance) and empower you to be part of the conversation," said event coordinator Senior Master Sgt. Amy, superintendent of the 526th IS, to the group of attendees. "Hopefully, as the concepts becomes less intimidating, they will stimulate a culture of curiosity within you that makes you want to learn more and dig deeper."

The question of whether you should let Facebook save your face is gaining in urgency as Facebook makes moves to expand its deployment of facial recognition. It faces a lawsuit by Illinois residents over the technology. SAN FRANCISCO -- Of all the information Facebook collects about you, nothing is more personal than your face. With 2.2 billion users uploading hundreds of millions of photos a day, the giant social network has developed one of the single-largest databases of faces and -- with so many images to train its facial recognition software -- one of the most accurate. The question of whether you should let Facebook save your face is gaining in urgency as it moves to expand its deployment of facial recognition, rolling it out in Europe, where it was scrapped in 2012 over privacy concerns and scanning and identifying more people in photos. At the same time, the giant social network is attempting to quash efforts to restrict the use of facial recognitionin the U.S., from legislation to litigation.

For anyone worrying about machines taking over the world, I have reassuring news: The idea of artificial intelligence has been overcome by hype. I don't mean to belittle AI's promise or even its existing capabilities. The technology allows organizations to put data to use in ways we could only imagine not that long ago. It's revolutionized the way executives approach strategic planning. But very often lately--when I'm in meetings, reading research papers or listening to an expert's presentation--I can't shake the feeling that to many people, terms like "AI," "machine learning" and "cognitive computing" have become answers unto themselves.