CommunityWatchdog.

Stop Unwanted Contents Before They Get Published.

Finally, Smart Analytics ! IntelligenceFocus’ CommunityWatchdog™ is a breakthrough online analytics on-demand solution which automatically analyzes and filters on-line content to enforce security compliance in real time for online social networks and Internet communities. Based on IntelligenceFocus’ proprietary Dynamic Intelligence Management System (DIMS), CommunityWatchdog permits community owners and managers to have their communities under control, monitoring in real time content compliance issues and proactively managing incoming risks with early warning capabilities.
Connectors Discovery Active ServicesOptimize Online User Experience by Filtering-out inappropriate and illegal content.

Now available in English, Danish, Dutch, Finnish, French, German, Italian, Norwegian, Spanish, Swedish, Portuguese, Turkish. CommunityWatchdog™ comprehensively mines unstructured and structured data flowing over a virtual community network without imposing complex software integration. CommunityWatchdog™ automatically analyzes and profiles the behavior of community users upon each and every upload and post event, understanding whether the user created content is compliant to the policies and rules of the community or national and international laws. CommunityWatchdog then manages all of the alerts captured by the Connectors and allows backend systems to block, delete or send alert messages.

Key Features

Examples

Global Scenario
Internet social networks and online communities for children are transforming the way children interact with one another. The virtual community is competing with traditional physical meeting places such as schoolyards, playgrounds, malls, etc.

This phenomenon has also given rise to new practices such as "friend hopping" where a community user finds someone that he or she knows, and, by looking at this person’s friends’ pictures and profiles, leads the user to “hop to” new friends by simply clicking on their pictures.

Young people are very computer savvy in this day and age, but even so, may not realize that posting information about themselves in an online community can expose their identities and where they live/work or play to people who potentially wish them harm. Even if the user him or herself doesn't post any personal information, sometimes friends’ postings can reveal such information.

Some of the emerging threats in Cyber Communities and Social Networks are:

• Identity Theft
• Exposure to Inappropriate Material
• Molestation, Harassment/Bullying
• Personal, Health and Financial Data Theft
• Privacy Infringement
• Copyrighted Material

Promoting Corporate and Government Online Social Responsability
Branding is moving beyond its concern with consumers alone, to become increasingly concerned with inspiring confidence among users and investors, creating a positive work environment for employees, and protecting the communities and the environment within which the enterprise operates. This shift to an holistic brand management strategy for attracting and retaining customers can benefit from building corporate reputation through the practice of corporate social responsibility. Potential damages and loss of value for unwanted content can be: negative impact on brand reputation, loss of online users and customers, lower ad revenue, lower conversion, lower loyalty, regulatory compliance issues, legal issues, class actions, financial loss.

How does it Work ?
IntelligenceFocus’ CommunityWatchdog™ is the first Active Service published by IntelligenceFocus. It uses the same backend system used by Discovery and BI Studio, and adds to the IntelligenceFocus suite the capability to modify live content and to perform a real time user-generated content matching with a predefined set of rules, patterns and taxonomies to prevent the publishing of inappropriate or illegal content.

CommunityWatchdog™ performs the following key actions

stop inappropriate content before they get published, and

report content compliance issues in the real time dashboard.

Artificial Intelligence based Content Matching Engine
CommunityWatchdog™ dynamically and intelligently monitors live contents being published in every type of inbound and outbound communication taking place within an online community. If certain content matches a known blacklist of keywords, URLs, conceptual patterns or inappropriate content specified within a certain taxonomy (e.g. sex related content), it is reported (and/or immediately blocked) in real time using a powerful and fully configurable security dashboard. Machine learning and pattern recognition ensure that CommunityWatchdog™ is easy to keep up-to-date. In effect, the system gets “smarter” the longer it is employed.

Ready
CommunityWatchdog™ is pre-configured to recognise the following type of text content: adult and sex related, violence, racism, profanity, offensive. Additionally it is possibile to easily extend rules, patterns and taxonomies to match virtually any kind of content and prevent this to be published.

Easy
CommunityWatchdog™ can be activated within any online web site, application, widget, game, etc. by simply inserting the right Connector for the programming language that is ised to create the web content. For example, a video game programmed in Microsoft C# and integrating an online community for kids and teens can easily be powered by CommunityWatchdog™ by few lines of .Net code; or an online Blog can use the Javascript connector for a seamless integration of CommunityWatchdog™ as a content filtering or intelligent moderation helper.

Multi Language
CommunityWatchdog™ works with over 12 languages including english, spanish, french, german, italian, etc. Asian languages will be available early in 2009. For a full list of supported languages please contact our tech support here.

nicknames, usernames

Prevent inappropriate nicknames and usernames to be published and exposed to online visitors.

reviews, comments

Identify and block badwords in free text across multiple languages and special internet slangs (eg. A = 4, E = 3 and proprietary soundlike-based textual algorithms).