Content discovery support

OpenTV Video Platform enables video content discovery for consumers. When integrated with ION, content discovery is exposed as a complete user experience flow or can be accessed at an API level for custom applications.

The key discovery journeys that are supported are:

• Search – live and VOD
• Electronic Program Guide
• Now & Next
• Structured VOD catalogue grouped by fields in the underlying metadata
• Editorial screen views, e.g., Magazine – combining manually selected items and recommendations into a single view

Content discovery enables integration of recommendations from XroadMedia Ncanto (and others by request).

XroadMedia Ncanto

Ncanto is a real-time content discovery and recommendation solution. It delivers highly personalized content via TVs, PCs, or mobile devices to extend viewer engagement and to maximize the revenue-generating potential of content assets. Content is sourced and targeted according to the service provider’s business requirements combined with the viewer’s profiles, viewing behaviour, and social graphs.

Ncanto provides the following benefits for the consumer:

• Discovers most relevant content across multiple content sources, helps to master the content overflow.
• Learns quickly (six times faster than competitor solutions), delivers relevant recommendations immediately, without training or observation periods and in real- time.
• Inherently identifies the usage context:
  • Does not need personal logon in case of shared usage (e.g., in the family/household).
  • Gives full control to share usage context across all devices and screen types.
• Dramatically reduces searching time and increases watching time.
• Delivers personal recommendations for multi-user environments without the necessity for them to login or identify themselves.
• Recommendations are always explained, helps the user to quickly build trust with the service.
• Increases consumer satisfaction.

Ncanto provides the following features for the service provider:

• Complete solution, comprising multiple content discovery and recommendation algorithms, made available through a single, standards-based API:
  • Search – full search functionalities to deliver either generic or personalised results, works across all content and asset sources
  • Editorial – moderated choices controlled by an editor
  • Similarity – e.g., using links between metadata describing a content (does not need user feedback or learning, requires minimum resources)
  • Collaborative – identifies other users with similar taste
  • Mood based recommendations and content filtering
  • Statistical – most watched, most liked, and other “best of” across the whole user base
  • Customer profile based – learns user taste and preferences from viewing behavior and user feedback as well as from the viewing settings
  • Social – learns user preferences from the user’s Facebook profile and graph (for example), as well as from the user’s own and followed tweets
  • Related online content – for easy discovery of relevant content on long-tail platforms (e.g., YouTube, Vimeo)
  • Viewing settings based – context-aware content discovery and recommendations (e.g., differs for the type of device, time of the day, day of the week, connectivity, etc.)
  • Any combination of the methodologies and algorithms above is configurable as well.
• Considers user feedback and changes in the asset store(s) immediately, always returns recommendations in real time.
• Supports filtering, sorting, and grouping of recommendation results based on any metadata field (e.g., genre, actor, HD, 4K, Dolby Surround, etc.) or user data characteristics (e.g., based on user segmentation).
• Considers the service provider’s and third party’s business goals (to increase revenues and profit, to exploit content rights to the maximum, to improve customer satisfaction and reduce churn, etc.) with the help of an integrated business logic.
• Supports multiple languages whereby the main language for a user can be automatically identified by Ncanto, set by the service provider and/or set by the user.
• Modular architecture – only the components relevant for a customer will be deployed.
• Scalable from tiny trial configurations to carrier-grade services with millions of subscribers (thousands of requests per second).
• Powerful monitoring and analytics functions to control the performance (response times, throughput) and the effectiveness (quality of recommendations, error rate, impact of business rules, A/B testing with user groups) of Ncanto.

Aggregation and deep linking

OpenTV Video Platform supports the ingest of third-party catalogues' metadata to enable the operator to show the additional titles alongside their own content. The third-party content can then be viewed using the OpenTV Player or within the third-party client depending on the agreement with the third-party:

• Operator has an agreement with a third-party to integrate their content.
• Metadata for their asset library is ingested into the OpenTV platform.
  • Metadata can be managed via the Operator Console.
• The assets are made available to clients via API.
  • The asset includes the direct links to content within the third-party application.
  • Or the stream URL if the third-party allows direct playing of the content.
• The OpenTV Baseline application displays the third-party content alongside the operator’s content in multiple locations.
  • This is achieved via manual curation.
  • Surfacing a third-party app's content using Google’s Channels API.
  • Or automated in the case of the virtual EPG channel.

Manual curation on the Home screen

As a virtual channel in the EPG

Shown in a strip on the  home menu (surfaced using the Google Channels API)

Inside the Media Card

• The OpenTV platform will then include the third-party content in its unified search results API, allowing the user to find the content they want to watch across services from within the OpenTV Platform.

Multiple services returned in the search results in the OpenTV Baseline Application