Media Mining System
Overview

Solutions & Features

The award-winning SAIL LABS Media Mining System is an integrated platform for analysts and decision makers that is able to extract metadata and key information in real time from multiple open sources even in multiple languages.

The system consists of multiple components that are combined to provide a revolutionary capability.

Overview

The system consists of multiple components; the Media Mining Feeder (MMF), the Media Mining Indexer (MMI),  the Text Mining Indexer (TMI), the Media Mining Server (MMS), the Media Mining Client (MMC), the Decision Maker User Interface (DMI) and the Crisis Room or Personal Situation Centre that work together to provide you a revolutionary new capability: the power of cross-media and cross-lingual information retrieval, analytics and dissemination.

The system automatically records and harvests data from multiple sources such as TV, Radio, YouTube, Internet (Social Media, Feeds, Websites), e-mails, and electronic documents. The raw data then runs through a series of processing steps: the speech content is transcribed, indexed and enriched with information regarding language, speaker, named entities, topics, and sentiment. It is subsequently stored on a database for later search and retrieval, and can also be archived. Visualisation and analysis of trends, relations, global hot spots, profiles, as well as ontologies and Social Media Analytics pave the way for turning raw, unstructured data into actionable knowledge and intelligence.

Media Mining Feeder

The Media Mining Feeder (MMF) for TV & FM Radio, MMF for Internet Video (YouTube), and MMF Internet Content and Email are SAIL LABS’ tool to continuously and concurrently capture and record unstructured data from multiple sources (digital satellite TV and Radio, FM Radio, YouTube, Websites/RSS & ATOM Feeds, Twitter, Facebook, Telegram, Vkontakte, Press Agencies and E-mails) and process it into structured, searchable and easily accessible information.

During the process, video and audio signals are transcoded and audio and meta-data are sent to the speech recognition engine, i.e. the Media Mining Indexer. The scheduled content is subsequently sent to the Media Mining Server for storage and can be accessed for search, retrieval and analytics through the Media Mining Client.

Media Mining Indexer

The Media Mining Indexer represents “the heart” of the Media Mining System. Here, speech recognition and other ground breaking technologies take place i.e. speech becomes text.

The Media Mining Indexer receives media streams through media feeders converting this input into a rich-transcript XML-file describing the content and metadata of the content. This rich-transcript is then sent to the Media Mining Server for storage, search and retrieval.

The Media Mining Indexer comprises a suite of powerful technologies for the processing of multimedia input: A subset of these work on the audio track of the input (Media Mining Indexer for Audio), some of them on the visual track (Media Mining Indexer for Video) others on the textual input or text generated by audio-or visual processing parts of the system (Media Mining Indexer for Text).

Media Mining Server

The Media Mining Server (MMS) is installed on two separate Hardware Devices. One hosts the Database, the other the Media Mining Client back end (Application Server). Together the Hardware Devices are the central hub of the Media Mining System, connecting the underlying technologies with the visualisation of the outer world  (i.e. Media Mining Client). Increasing volume of cross-media data requires means for storage and easy access to information.
The Media Mining Server turns the audio and video library from a collection into a tool, accessible for search, retrieval and analytics through the Media Mining Client (GUI).

Media Mining Client

The Media Mining Client (MMC) is the User Interface to the Media Mining Server and designed for search, retrieval and analytics of relevant content. Furthermore, it enables filtering vast amounts of cross-media and cross-lingual content and monitoring the media regarding issues, regions and persons through visualisation of trends, relations, clusters, ontologies and geographical relations as well as usage of workflow features.

In addition, the Social Media Extension (SME) provides new meta data collection and analytics as well as visualisations of various Social Media analysis aspects and is fully integrated into the Media Mining Client (MMC).

Decision Maker Interface

The Decision Maker Interface (DMI) is a stand-alone application designed to be operated on various devices (PC, tablet, handheld) as well as print-out, enabling the decision maker to automatically receive all approved notifications (clips, OSInfIs, profiles or reports). Decision makers get a quick overview of topics of interest and – if needed – a detailed understanding of the previously performed steps by tracing the source data.

Crisis Room
&
Personal Situation Centre

The Media Mining Crisis-Room and Media Mining Personal Situation Centre for TV Screens is a system dedicated to enhance decision-support of management and analysts in crisis situations (i.e. financial crisis, natural disasters, industrial accidents, targeted acts of violence, etc.). It enables simultaneous monitoring of live media-streams across multiple channels in various languages. During this monitoring process, users can perform ‘live searches’, i.e. search for a specific keyword across any number of live channels without watching the original video. The system collects the results automatically and alerts users as new matches to their search criteria occur. The number and type of monitored channels can be selected and arranged freely according to the user’s needs. They can be displayed on industry-standard TV Screens.

Close Menu