The Cool Bits

The data model used to expose the metadata to back-end storage mechanisms is critical to searching, browsing, and otherwise exploiting the information. Autonomy Virage’s software watches, listens, reads, and allows annotation of video. As metadata is squeezed out of the video signal, it is placed into a data track where each data element is time-synchronized with the corresponding video representations. Autonomy Virage offers a selection of Media Analysis Plug-ins to analyze the different components in the video file to extract all the relevant information from deep within the video, creating a uniquely granular index.

Hyperlinking
Autonomy Virage’s software allows manual and fully automatic linking between related pieces of information regardless of their format. The concepts in a document can be linked automatically to those in another file, connecting information at an unprecedented level of granularity. Video clips do not need to be broken down into segments and dealt with separately; rather, the concepts within a file can be extracted and linked to the specific section of another video where that same concept is discussed. This allows the video to be stored in its original format while retaining the benefits usually associated with working with smaller clips.

Hyperlinks are generated in real-time at the moment a document is viewed, removing the need for any manual intervention and ensuring they are constantly up-to-date. Unlike other approaches, these hyperlinks aren’t simply based on user-generated metatags, but rather on an understanding of the video file, including the images shown and the audio track. This is a powerful and valuable tool for research and also far more accurate than traditional, human based techniques. BBC journalists use Autonomy technology during the editorial process to accelerate newsroom workflow and provide additional content that they can push out to their readers automatically and in real-time.

Sentiment Analysis
By analyzing the structures and meaning of language, IDOL determines the positive and negative characteristics of each piece of information and creates relevant classification systems. IDOL can determine the degree to which a sentiment is positive, negative or neutral for the entire content or a segment of the content. In addition, administrators can apply multiple tagging functions and specific threshold cut-offs to determine the sensitivity of sentiment analysis.

Cloud Analysis
Autonomy is able to identify the main concepts discussed within a set of data and present them to users as an idea cloud. The Cloud contains is a list of terms that are conceptually represented in the search results. Clicking on the term will result in a new subset of information result based on that idea.

Entity Extraction
Autonomy Virage’s unrivalled portfolio of connectors and plug-ins allow IDOL to recognize information and automatically extract it from within the file. By making it possible to recognize items within the video file, such as the Statue of Liberty or other famous landmarks, Autonomy Virage is able to extract the probable location of the scene. IDOL is able to enrich extracted data, based on the knowledge held within the organization, automatically cross-referencing new information with old in order to determine which entities have changed. Furthermore, the technology can also extract other entities such as names and logos making it a powerful brand and intellectual property monitoring tool, for example, automatic identification of product placement within a film.

Social Media Analysis
Every day, the social capabilities of the web continue to increase in size and influence with consumers. Social networking sites like Facebook, IMDb, LinkedIn, Twitter, and YouTube represent an ever-expanding forum for expressing sentiment, preferences, and trends. At the same time, marketers are grappling with how to harness the chaotic and constantly shape-shifting nature of the web, as traditional analytics and monitoring solutions are ill-equipped to access and understand the unstructured information that lives on social networks.

Powered by IDOL, Autonomy’s Social Media Analysis is an industry-first offering to provide marketers the technology needed for effective brand measurement and reputation management. It provides the ability to create and manage communities with out-of-the- box social media functionality, listen and understand the exploding amount of user-generated content on the web, and proactively leverage and act on online dialogue.

Cognitive
Due to the unique combination of Bayesian Inference and Shannon’s Information Theory at the core of the technology, Autonomy software is able to continuously develop and learn. This learning ability significantly reduces the manual input required by other solutions and translates into large savings in time and money for our clients. Where other solutions need to be taught new words, phrases or concepts and shown how to categorize them, Autonomy can automatically deduce the significance of these new units of meaning, add them to relevant categories or create new ones where necessary.

Subscribing and Alerting
By understanding who needs or wants what kind of information, IDOL can deliver it as it becomes available. This is particularly crucial in broadcast monitoring operations where multiple channels are being monitored simultaneously and users must react quickly to new developments.

Stay ahead: Automatic profiling and alerting enables journalists to stay in touch with breaking news and significant events in order to accelerate the reporting and publishing process.

IDOL delivers personalized content through various channels, including browsers, mobile devices (phones, PDAs, etc.) and even digital TV. Not only can the same content be repurposed for each device, but user interaction through any device also helps IDOL understand user interests and expertise.

Key Benefits:

• Provides highly relevant and accurate information to be delivered to an individual automatically
• Allows informed decisions to be made more quickly
• Makes information widely accessible
• Increases competitive edge

Hot and Breaking Topics
Keeping a real-time pulse on events enables organizations to pro-actively address issues or opportunities, creating a competitive and market advantage. Autonomy Virage can automatically analyze all data types, including news stories, broadcast streams, instant messaging, and other sources to determine "Hot" or "Breaking" trends within the content. "Hot Topics" are the highest volume of topics discussed, and "Breaking Topics" identifies new trends or subjects. These can then be dynamically displayed to the user.

Time Variance
Information regarding developing stories and on-going subjects can be monitored and displayed over time on a Spectrograph pane where the information is arranged into self-similar groups, or clusters. The x-axis represents information over time, enabling users to visualize how clusters develop over a given time period; the y-axis represents the range of concepts defined within the knowledge base. The spectrograph can display both “hot news” and “breaking news” in the same instance. By scrolling over the spectrograph, titles are automatically generated and assigned to every cluster. By clicking on the cluster, the results can be viewed. The importance of clusters over time can be seen through the change of color and width. The color and intensity of the lines are indications of the size of a cluster. The brighter colors indicate “what’s hot” while the width of the lines is an indication of the volume of information on that topic.

Cluster Mapping
Autonomy Virage’s cluster mapping automatically identifies inherent themes found within audio and video files, and then groups the similar concepts together in a visual format. By displaying the information in a two dimensional cluster map, organizations are able to quickly ascertain similarities and differences. Clusters rank search results by conceptual similarity, which is essential to locating interactions that share the same meaning, even when they contain different keywords.

Media Rendering
Autonomy Virage provides a unique capability to store the master high resolution “core asset” and create derivatives on-thefly based on the selection and desire of the requestor. The transformation engine supports all of the most common rich media formats, and utilizes a multi-threaded task engine capable of handling large quantities of simultaneous complex transformations involving format conversions, color-space conversions, color adjustments, resolution, cropping, sizing, padding, watermarking and a wide variety of advanced graphics adjustments that would normally require a user to open an editing application on their desktop.

Cross-Lingual
Autonomy’s technology uses probabilistic modeling to extract meaning from digital content and forgoes language-dependent parsing or dictionaries to form ideas. Because Autonomy treats words merely as abstract symbols of meaning, it is completely language independent. It does not rely on an intimate knowledge of a language’s grammatical structure, but rather derives its understanding through the context of the words’ occurrence rather than through a rigid definition of grammar. This highly mathematical logic yields high accuracy, and performance is further optimized through proprietary stemming algorithms, “sentence breaking” libraries, stoplists and n-grams.