ContentAssess: Automatic assessment of article quality
Project Description Market Need • In the modern online media landscape, there are often a wide range of articles from different media sources covering the same topic. • For busy web users who wish to obtain a quick view on a topic, it can be difficult to evaluate the best article to read. Technology Solution We have developed a number of measures spread over three complementary dimensions for automatically assessing article quality. Authority: This dimension takes into account the reputation of the source of the article along with its level of domain expertise and specificity. Social Signal: Multiple measures derived from
StreamConverge: Spark streaming over heterogeneous datasets
Project Description Market Need • Complex systems often include many different streaming data formats produced by different system outputs, sensors and sub-systems. • Often there is a need to integrate these homogeneous data streams in real-time in an efficient, scalable and fault tolerant way. Technology Solution We have implemented a heterogeneous streaming data integration demonstrator on Apache Spark Streaming which is: • Tolerant to out-of-order event arrival • Merges and integrates events by key • Scalable and efficient for high velocity, high volume data Applicability • The technology can be used across many different scenarios where heterogeneous streaming data is
ContactFiller: Extracting structured entity profiles from unstructured sources
Project Description MARKET NEED Information is key in the modern commercial landscape and often we have very little time to do due diligence about a prospective client or business contact before or after a meeting. There is a need for a rapid information extraction system that takes a set of information about a person or business and searches the Web and other unstructured data sources for more specific information about that entity. While there are existing solutions that allow business cards to be scanned and the information contained in them to be extracted, there are not tools that fully exploit
ContinuousMetrics: Real-time metrics computation for continuous streaming data
Project Description MARKET NEED Computing metrics in real-time over live streaming data sources is required in many industries. Financial services companies compute metrics within a sliding window for analysing stock prices, trading volumes and fraud detection. Energy suppliers analyse live power usage and grid statistics. Telecommunications companies analyse real- time network usage and health statistics. Transportation and logistics agencies provide real-time traffic information to drivers based on live metrics computed over traffic flows at key road intersections. In many scenarios these metrics must often be computed as fast as possible (low latency) over high volume, high velocity data streams (high throughput).
Social Identity Fingerprint: Matching user profiles across social networks
Project Description MARKET NEED There is great interest from many companies in matching user profiles across diverse social networks. Users use different social networks in different ways and it can be a challenging task to match the user activity across these networks, especially if profile information is missing or obfuscated in some way There is also strong interest in finding users with similar traits (connections, content, names, activity) within the same social network and the SocialIdentityFingerprint system allows the user to apply the cross-network metrics also within a network. TECHNOLOGY SOLUTION We have developed a web-based search tool which allows
LookAndLearn: Continuous data stream analytics for images
Project Description Market Need Object detection, especially small object detection, is a well known and studied image analysis challenge. An example of which is finding predefined brands/logos in a large number of images or a continuous image stream in real-time. Current logo detection applications are generally run in an offline way, and cannot handle a large, continuous image data stream and process it in real time or near real-time. Technology Solution CeADAR’s LookAndLearn project invents a technique for recognizing brands/logos within high-throughput image data streams. For example, a stream of digital photographs or video frames may contain recognizable visual brand
Forecasting Technology Platform: A toolbox of advanced analytics predicting the future using past data streams
Project Description Market Need Different areas of the economy require precise forecasts of future events based on knowledge collected from the past. Technology platforms which are accurate and effective in acquiring and processing such information are of great value to the market because they assist in planning and helping to prepare for what is coming. Unlike typical forecasting which can be based on informal methods, optimal forecasting methodology can be conducted based on scientific tools to assure proper outputs. As a result, appropriate feedback on a given problem can successfully warn if the process under investigation is heading towards an
IdentityMatch: Identifying persons of interest in social network contexts
Project Description Market Need Finding persons with certain behaviours and attributes in network data is a challenging task. Current approaches may only take into account content information and discard any network influences and metrics Growing need for people aggregator systems to locate users with particular interests or skillsets Technology Solution We use statistical text analytics, and network analysis to build a network of related users. We then refine this network based on a set of topics created by the user plus reputation and social influence metrics. User-defined topic keywords are augmented from the live Twitter stream to enrich the search
Advanced monitoring of entities in online media sources: Online monitoring of the reputation of entities through social media
Project Description Market Need In the always – connected 24 – 7 online marketplace, up – to – date knowledge is paramount for good decision making . In many timezones across the globe, content is emerging which can have an effect on the reputation of a firm, individual or product . Depending on the trust level attached to the source, a single social media message can cause significant damage to the reputation of an entity, as experienced when a tweet was sent from the hacked Associated Press Twitter account, causing th e Dow Jones Index on Wall Street to drop