Real-time search is a challenge due to the sheer volume of information and the difficulty to filter and find relevant information.
Filtize (www.filtize.com) is a start up working on new ways to improve access to real time information.
The service improves real time Twitter search through data processing, filtering and enhanced visualization and navigation.
It can be used by media organizations who need to track searches and find specific information or marketing companies looking for trends information.
The service is supported by innovative technology and functionality:
. The service allows to define persistent searches that are constantly running on the server and for which data is retrieved in real-time from Twitter and cached on Filtize server. The service also has defined a few domain specific searches.
. The data is processed and enriched on Filtize server using natural language processing technologies including data categorization, filtering, content and topic matching, statistical analysis, etc.
The service can work with any language but is currently optimized for English, Spanish and French languages.
. A rich user web interface allows to navigate and filter the searches and visualize additional information including meta data (whether a link points to text, sound or video content for instance), trends and sentiment analysis.
. The UI also allows to reuse Twitter web interface to look at specific users for instance.
. The service offers a scalable and modular platform architecture to quickly and easily expand features and capabilities
I am working on enhancing the service which will include additional capabilities around data curation, additional filtering, data clustering and alerting.
I am welcoming suggestions, feedback and funding propositions.