Quality of Information-Aware Resource Allocation: In-Network Processing
- This demonstration showed two video streams transmitted from a source to two different destinations. The capacity of the wireless network will not support the video streams at the full native rate. If no controls are placed in the network, the video quality at the receivers will be very poor. Even with standard Quality of Service methods applied through scheduling algorithms like Fair Queuing or Weighted Fair Queuing, the video quality will be very poor for at least one recipient because of the highly constrained wireless links. This illustrates that using only controls and knowledge of the communications network is insufficient for supporting the missions in this case. When Quality of Information-Aware algorithms are applied, we consider that the information derived from the videos may still be obtained if the videos are compressed in a controlled fashion in the network. This allows both receivers to receive the videos at sufficient quality for their purpose. The demonstration shows the last case. The QoI-ware controls exercise two important genres of network. First, from information networks we determine the limits of in-network processing, in this case compression that can be done to transform our source data into data that meets QoI requirements. Second, resource allocation is done on the communication network to ensure that the data is delivered according to desired metrics so that the necessary information may be extracted. The poster shows how resource allocation would be done with no controls, with scheduling mechanisms, and with the QoI-Aware controls.
Controllability of Complex Networks
- Controllability is one of the central concepts in control theory. It addresses our ability to drive a system from any initial state to any final state in finite time. Here, we apply control theory to complex networks and show a simple method to identify the minimum set of driver nodes we need to control an arbitrary directed weighted network.
TopicLens: Exploring Topic Related Trust in a Network User Interface
- Visually communicating credibility and trust information is a challenging problem. This demonstration will focus on TopicLens, an interactive system supporting analysis of probabilistic associations in topic-modeled data in conjunction with the structure in the underlying social network from which the modeled documents were sourced. The design consists of a hybrid view, containing a river-like representation to communicate credibility or trust that a user (or group of users) have with respect to a given topic. These factors can be assessed though an examination of a users probabilistic association with that topic, and also through examination of that user’s social network and their respective credibility within the selected topic. The design leverages interactive input mechanisms such as hovering, selection and rotation to generate informative and intuitive outputs, which can take the form of label and item highlighting, opacity variance, or connection lines, all contributing to a better understanding of topic-based trust and expertise across the network.
Discovering Factors of Trust in Social Networks
- A task of primary importance for social network users who are interested in receiving information about particular topics is to decide which other users’ updates to subscribe to in order to maximize the relevance, credibility, and quality of the information received. To address this problem, we conducted an experiment designed to measure the extent to which different factors in online social networks affect both explicit and implicit judgments of credibility. The results of the study indicate that both the topical content of information sources and social network structure affect source credibility. Based on these results, we designed a novel method of automatically identifying and ranking social network users according to their relevance and expertise for a given topic. We performed empirical studies to compare a variety of alternative ranking algorithms and a proprietary service provided by a commercial website specifically designed for the same purpose. Our findings show a great potential for automatically identifying and ranking credible users for any given topic.
Construction, Search, Mining of Self-Boosting Text-Rich Information Networks
- Approximately 80% percent of information is held in an unstructured format. For example, thousands of “attack” events and hundreds of “arrest” events can be mined from one week’s unstructured textual data, such as blogs, twitters, news and reports. We will demonstrate a system which can automatically extract multi-dimensional information networks from unstructured data, containing rich facts about entities, relations and events that may facilitate a military analyst in terrorist information search gathering, control and analysis for any given queries. This demonstration also shows how the quality of information can be improved by the interconnected network itself, which we call self-boosting of information networks.
Apollo: Fact Finding from Noisy Data
- Open information sources, such as the Twitter stream, while potentially providing eye-witness accounts of developing crises and events, are also notoriously full of incorrect facts and irrelevant blather. This technology filters and integrates heterogeneous types of information by weighing input from multi-genre sensor, social, and information networks, allowing users to quickly skim only the most credible and relevant content and providing a tool for truth and trust discovery within a target area.
Link Prediction: Will and when a link or relationship emerge in the future?
- Complex networks are always evolving, with links forming and breaking. This approach shows how link formation can be predicted and the effects on the overall network visualized, whether the network consists of social, information, or communications connections.
Multi-Team Systems Simulation
- Whether it’s emergency relief due to natural disasters or humanitarian aid in war torn regions, there are situations where international organizations, first responders, and military personnel need to collaborate effectively on teams in stressful situations. Researchers at Northwestern University and the University of Illinois at Urbana Champaign will use the Multi-team System Simulation – or MTS Platform – to gain insight on how network parameters can be configured to better allow small teams to coordinate in such emergency situations.
NS CTA Composite Experimentation Infrastructure and Methodology
- This video will give an overview of the experiments the R1 group has put together. Our overall goal is to provide militarily relevant experiments in composite network environments with the purpose of providing new insight into NS CTA research as well as producing transition opportunities. Because of trade-offs inherent in complexity, our methodology is to conduct experiments at several levels of fidelity. This allows us to observe behaviors in more simple, highly controlled scenarios as well as behaviors in more complex scenarios that emulate real world phenomena.