Traditional approaches using Deep Neural Networks for classification, while unquestionably successful, struggle with more general intelligence tasks such as "on the fly" learning as demonstrated by biological systems. Organisms possess myriad sensory organs for interacting with their environment. By the time these diverse sensory signals reach the brain; however, they are all converted into a spiking information representation, over which the brain itself operates. In a similar manner, myriad machine learning (ML) algorithms today compute on equally diverse data modalities; but without a consistent information representation for their respective outputs, these algorithms are frequently used independently of each other. Consequently, there is growing interest in information representations to unify these algorithms, with the larger goal of designing ML modules that may be arbitrarily arranged to solve larger-scale ML problems, analogous to digital circuit design today. One promising information representation is that of a "symbol" expressed as a high-dimensional vector, thousands of elements long. Hyperdimensional computing (HDC) is an algebra for the creation, manipulation, and measurement of correlations among "symbols" expressed as hypervectors. Towards this goal, an external plexiform layer (EPL) network, echo state network (ESN), and modern Hopfield network were adapted to implement the mathematical operations of complex phasor based HDC. Further, since symbol error correction is an important consideration for computing with networks of ML modules, a task agnostic minimum query similarity for complete symbol error correction was measured as a function of hypervector length. Based on these results, problem-independent similarities have been established within which HDC equations should be designed. Lastly, these ANNs were tested against several tasks representative of online and "plug & play" ML among expeditionary robots. For all criteria considered, the modern Hopfield network was the most capable ANN evaluated for use with complex phasor based HDC, providing 100% symbol recovery in a single time step for nearly all parameter settings.

During the 2016 SPIE DSS conference, nine panelists were invited to highlight the trends and opportunities in cyber-physical systems (CPS) and Internet of Things (IoT) with information fusion. The world will be ubiquitously outfitted with many sensors to support our daily living thorough the Internet of Things (IoT), manage infrastructure developments with cyber-physical systems (CPS), as well as provide communication through networked information fusion technology over the internet (NIFTI). This paper summarizes the panel discussions on opportunities of information fusion to the growing trends in CPS and IoT. The summary includes the concepts and areas where information supports these CPS/IoT which includes situation awareness, transportation, and smart grids.

During the SPIE 2012 conference, panelists convened to discuss “Real world issues and challenges in Human Social/Cultural/Behavioral modeling with Applications to Information Fusion.” Each panelist presented their current trends and issues. The panel had agreement on advanced situation modeling, working with users for situation awareness and sense-making, and HSCB context modeling in focusing research activities. Each panelist added different perspectives based on the domain of interest such as physical, cyber, and social attacks from which estimates and projections can be forecasted. Also, additional techniques were addressed such as interest graphs, network modeling, and variable length Markov Models. This paper summarizes the panelists discussions to highlight the common themes and the related contrasting approaches to the domains in which HSCB applies to information fusion applications.

Situation Assessment (SA) modeling has many instances of development, but there has yet to be a comprehensive set of metrics used for performance evaluation. The amount of data being presented and displayed to the analyst is overwhelming — to a point that in many cases they are missing the salient or key activities of interest (AOI). Analysts are spending the majority of their time filtering through the data rather than performing analysis. To aid the user (high-level information fusion), we explore the nature of how we can rank various activities based on their impact and threat. We develop a information fusion SA reference model over SA metrics of confidence (precision, recall), accuracy, activities of interest, timeliness and throughput to determine the intent and assessment of situation activities.

: This proposal addresses an emerging challenge of developing a computational understanding of online groups and their interactions. By investigating the interactions between the groups and the social, economical, and physical forces acting on them, this project endeavors to achieve a better understanding of the operational environment by extracting pertinent values from groups of special interests. Online data sources contain massive amounts of data. The state-of-art search engines are designed to help general query-specific search and not suitable for finding disconnected online groups. The scientific and technical merits of the proposed research are demonstrated in (1) formulating novel research problems of searching disconnected online groups, (2) developing innovative mathematical and statistical models and efficient algorithms that leverage existing search engines and employ contextual information and meta data, and (3) conducting interdisciplinary scientific research that enables a new kind of search power and complementary research capabilities.

The National Operational Environment Model (NOEM) is a strategic analysis/assessment tool that provides insight into the complex state space (as a system) that is today's modern operational environment. The NOEM supports baseline forecasts by generating plausible futures based on the current state. It supports what-if analysis by forecasting ramifications of potential "Blue" actions on the environment. The NOEM also supports sensitivity analysis by identifying possible pressure (leverage) points in support of the Commander that resolves forecasted instabilities, and by ranking sensitivities in a list for each leverage point and response. The NOEM can be used to assist Decision Makers, Analysts and Researchers with understanding the inter-workings of a region or nation state, the consequences of implementing specific policies, and the ability to plug in new operational environment theories/models as they mature. The NOEM is built upon an open-source, license-free set of capabilities, and aims to provide support for pluggable modules that make up a given model. The NOEM currently has an extensive number of modules (e.g. economic, security & social well-being pieces such as critical infrastructure) completed along with a number of tools to exercise them. The focus this year is on modeling the social and behavioral aspects of a populace within their environment, primarily the formation of various interest groups, their beliefs, their requirements, their grievances, their affinities, and the likelihood of a wide range of their actions, depending on their perceived level of security and happiness. As such, several research efforts are currently underway to model human behavior from a group perspective, in the pursuit of eventual integration and balance of populace needs/demands within their respective operational environment and the capacity to meet those demands. In this paper we will provide an overview of the NOEM, the need for and a description of its main components. We will also provide a detailed discussion of the model and sample use cases.

This book constitutes the refereed proceedings of the 4th International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction, held in College Park, MD, USA, March 29-31, 2011. The 48 papers and 3 keynotes presented in this volume were carefully reviewed and selected from 88 submissions. The papers cover a wide range of topics including social network analysis; modeling; machine learning and data mining; social behaviors; public health; cultural aspects; and, effects and search.

Abstract : Many say we live in the information age, but in reality if you ask any analyst today they would say we live in the data age. The amount of data being presented and displayed to the analyst is overwhelming - to a point that in many cases they are missing the salient of key activities of interest. Analysts are spending the majority of their time filtering through the data rather than performing analysis. Until recently, in the past five years, has there been an increased emphasis in higher level fusion or what many are calling situation awareness. So why the increased interest? In this paper we will look at this very issue, review our reference model and provide a discussion of a flow through the model to include how we can rank various activities based on their impact and threat.

In this work, we examine familiar strangers on Blogosphere and issues of finding them. In our daily life, familiar strangers, as coined by Stanley Milgram, do not know each other, but frequently exhibit some common patterns. Blogosphere is a part of the Web where bloggers post in individual or community blog sites. The nature of the Web is a scale-free network, which determines that a power law distribution applies to bloggers. That is, the majority bloggers are only connected with a small number of fellow bloggers, and these blogging groups are largely disconnected from each other. Familiar strangers on Blogosphere are not directly connected, but share some patterns in their blogging activities. We present a new problem: Aggregating familiar strangers on Blogosphere that allows for better personalized services, targeted marketing, exploration of new business opportunities, and predictive modeling and marketing. Finding familiar strangers on Blogosphere presents a challenge resulting from their disconnectedness. We look at typical blogs and understand the status quo, while seeking innovative ways to improve business intelligence. We define the problem of searching for familiar strangers on Blogosphere, elucidate the significance of doing so, study the challenges of finding them, and present and discuss some potential approaches. 1. Familiar Strangers on Blogosphere The advent of Web 2.0 [2] has started a surge of opensource intelligence via online media such as blogs. Since more and more people are participating in Web 2.0 activities, it has generated enormous amounts of collective wisdom or open-source intelligence. Web 2.0 has allowed the mass not only to contribute and edit posts/articles through blogs and wikis, but also enrich the existing content by providing tags or labels, hence turning the former information consumers to the new producers. Allowing the mass to contribute or edit has also increased collaboration among the people unlike Web 1.0 where the access to the content was limited to a chosen few. Blogs are invigorating this process by encouraging the mass to document their ideas, thoughts, opinions, views reverse chronologically, called blog posts, and share them with other bloggers. These blog posts are published on blog sites and the universe of these blog sites is called Blogosphere. Familiar strangers on Blogosphere are not directly connected, but share some patterns in their blogging activities. Figure 1: Familiar Stranger Bloggers and Long Tail

The nature of the Blogosphere determines that the majority of bloggers are only connected with a small number of fellow bloggers, and similar bloggers can be largely disconnected from each other. Aggregating them allows for cost-effective personalized services, targeted marketing, and exploration of new business opportunities. As most bloggers have only a small number of adjacent bloggers, the problem of aggregating similar bloggers presents challenges that demand novel algorithms of connecting the non-adjacent due to the fragmented distributions of bloggers. In this work, we define the problem, delineate its challenges, and present an approach that uses innovative ways to employ contextual information and collective wisdom to aggregate similar bloggers. A real-world blog directory is used for experiments. We demonstrate the efficacy of our approach, report findings, and discuss related issues and future work.

introduction Share on Introduction to special issue on social computing, behavioral modeling, and prediction Editors: Huan Liu View Profile , John Salerno View Profile , Michael Young View Profile , Rakesh Agrawal View Profile , Philip S. Yu View Profile Authors Info & Claims ACM Transactions on Knowledge Discovery from DataVolume 3Issue 2April 2009 Article No.: 6pp 1–3https://doi.org/10.1145/1514888.1514889Published:21 April 2009Publication History 6citation751DownloadsMetricsTotal Citations6Total Downloads751Last 12 Months2Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

The pervasive use of computer and Internet technologies creates an unprecedented environment where people can share opinions and experiences, exchange ideas, offer suggestions and advice, debate and even conduct experiments. Social computing, the study of social behavior and context based on computational systems, facilitates behavioral modeling in model building, analysis, pattern mining, anticipation, and prediction. This unique volume presents material from the second interdisciplinary workshop focused on employing social computing for behavioral modeling and prediction. The book provides a platform for disseminating results and developing new concepts and methodologies aimed at advancing and deepening our understanding of social and behavioral computing to aid critical decision making. The contributions incorporate views from government, industry and academia, and address research problems arising from pressing demands in the real world.

Information fusion system designs require sensor and resource management (SM) for effective and efficient data collection, processing, and dissemination. Common Level 4 fusion sensor management (or process refinement) inter-relations with target tracking and identification (Level 1 fusion) have been detailed in the literature. At the ISIF Fusion Conference, a panel discussion was held to examine the contemporary issues and challenges pertaining to the interaction between SM and situation and threat assessment (Level 2/3 fusion). This summarizes the key tenants of the invited panel experts. The common themes were: (1) Addressing the user in system control, (2) Determining a standard set of metrics, (3) Evaluating fusion systems to deliver timely information needs, (4) Dynamic updating for planning mission time-horizons, (5) Joint optimization of objective functions at all levels, (6) L2/3 situation entity definitions for knowledge discovery, modeling, and information projection, and (7) Addressing constraints for resource planning and scheduling.