IBM announced that its Watson AI is getting image recognition capabilities earlier in the year, but this site that lets you feed in your own photos to see what it thinks is in them is both impressive and scary. The visual recognition demo lets you give Watson an image URL or upload a photo and it'll come back in a few seconds with what it thinks it sees. This year's edition of TNW Conference in Amsterdam includes some of the biggest names in tech. In my tests I fed Watson a few random photos I had on hand and the accuracy was quite surprising. It could figure out what was in landscape shots, animals (down to the breed) and even what's in the background.

Image recognition is a hot area of research using artificial intelligence, and now IBM offers an online demo to let anyone test out the capabilities offered by its Watson cognitive computing system. Six sample photos are provided for illustration, or you can upload your own and ask Watson to analyze them. Either way, the cognitive system will produce a series of "classifiers" offering descriptions of the image's contents along with confidence scores for each of them. You can also create custom classifiers tailored for specific purposes. Watson gained worldwide fame when it won on the quiz show Jeopardy back in 2011, and IBM has been developing commercial applications ever since.

IBM has updated its artificial intelligence (AI) product, IBM Watson, giving it the ability to recognise images. Watson, which relies on cognitive learning to help it process the world in a human-like manner, can now'guess' what's happening in images fed to it via URLs. IBM has created a'Visual Recognition Demo' to showcase Watson's latest trick, which allows users to feed Watson an image before it tells you what it believes it sees. For example, supplying Watson with the image of a tiger throws up the result 77 per cent tiger, 26 per cent wild cat and 63 per cent cat. As well as identifying objects, people or animals in photos, Watson is also fairly adept at guessing what's going on in the background of images such as sunsets and other outdoor scenes.

Last month I wrote an article describing the interfaces and capabilities of Microsoft and IBM's new cloud data science products. I observed that Azure ML presents a user-friendly drag and drop data mining app for businesses, while Watson Analytics focuses on natural language queries but is still too nascent for use. A similar query for "IBM Watson Analytics" turns up 730,000 documents. Amid the deluge of coverage on both services, one could lose sight of the many upstart companies offering cloud machine learning services. However, new product categories are typically pioneered by startups.

One long-term goal of machine learning research is to produce methods that are applicable to reasoning and natural language, in particular building an intelligent dialogue agent. To measure progress towards that goal, we argue for the usefulness of a set of proxy tasks that evaluate reading comprehension via question answering. Our tasks measure understanding in several ways: whether a system is able to answer questions via chaining facts, simple induction, deduction and many more. The tasks are designed to be prerequisites for any system that aims to be capable of conversing with a human. We believe many existing learning systems can currently not solve them, and hence our aim is to classify these tasks into skill sets, so that researchers can identify (and then rectify) the failings of their systems. We also extend and improve the recently introduced Memory Networks model, and show it is able to solve some, but not all, of the tasks.

For the vast majority of queries (for example, navigation, simple fact lookup, and others), search engines do extremely well. Their ability to quickly provide answers to queries is a remarkable testament to the power of many of the fundamental methods of AI. They also highlight many of the issues that are common to sophisticated AI question-answering systems. It has become clear that people think of search programs in ways that are very different from traditional information sources. Rapid and ready-at-hand access, depth of processing, and the way they enable people to offload some ordinary memory tasks suggest that search engines have become more of a cognitive amplifier than a simple repository or front-end to the Internet. Like all sophisticated tools, people still need to learn how to use them. Although search engines are superb at finding and presenting information—up to and including extracting complex relations and making simple inferences—knowing how to frame questions and evaluate their results for accuracy and credibility remains an ongoing challenge. Some questions are still deep and complex, and still require knowledge on the part of the search user to work through to a successful answer. And the fact that the underlying information content, user interfaces, and capabilities are all in a continual state of change means that searchers need to continually update their knowledge of what these programs can (and cannot) do.

This work aims to address the problem of image-based question-answering (QA) with new models and datasets. In our work, we propose to use neural networks and visual semantic embeddings, without intermediate stages such as object detection and image segmentation, to predict answers to simple questions about images. Our model performs 1.8 times better than the only published results on an existing image QA dataset. We also present a question generation algorithm that converts image descriptions, which are widely available, into QA form. We used this algorithm to produce an order-of-magnitude larger dataset, with more evenly distributed answers. A suite of baseline results on this new dataset are also presented.

This work aims to address the problem of image-based question-answering (QA) with new models and datasets. In our work, we propose to use neural networks and visual semantic embeddings, without intermediate stages such as object detection and image segmentation, to predict answers to simple questions about images. Our model performs 1.8 times better than the only published results on an existing image QA dataset. We also present a question generation algorithm that converts image descriptions, which are widely available, into QA form. We used this algorithm to produce an order-of-magnitude larger dataset, with more evenly distributed answers. A suite of baseline results on this new dataset are also presented.

Community Question Answering (CQA) websites are a popular tool for internet users to fulfill diverse information needs. Posted questions can be multiple sentences long and span diverse domains. They go beyond factoid questions and can be conversational, opinion-seeking and experiential questions, that might have multiple, potentially conflicting, useful answers from different users. In this paper, we describe a large-scale formative study to collect commonsense properties of questions and answers from 18 diverse communities from stackexchange.com. We collected 50,000 human judgments on 500 question-answer pairs. Commonsense properties are features that humans can extract and characterize reliably by using their commonsense knowledge and native language skills, and no special domain expertise is assumed. We report results and suggestions for designing human computation tasks for collecting commonsense semantic judgments.

How should we gather information in a network, where each node's visibility is limited to its local neighborhood? This problem arises in numerous real-world applications, such as surveying and task routing in social networks, team formation in collaborative networks and experimental design with dependency constraints. Often the informativeness of a set of nodes can be quantified via a submodular utility function. Existing approaches for submodular optimization, however, require that the set of all nodes that can be selected is known ahead of time, which is often unrealistic. In contrast, we propose a novel model where we start our exploration from an initial node, and new nodes become visible and available for selection only once one of their neighbors has been chosen. We then present a general algorithm \elgreedy for this problem, and provide theoretical bounds on its performance dependent on structural properties of the underlying network. We evaluate our methodology on various simulated problem instances as well as on data collected from social question answering system deployed within a large enterprise.