In the previous posts we were talking mostly about data – how to harvest, index,process, manage and display information from multiple sources including blogs, forums, social hubs , news portals, premium content providers etc. What we didn’t mentioned is that the information always comes complemented with extremely useful metadata and is a matter of a smart technology to extract the most valuable pieces out of it.
Why is meta data so important? First, there are some important things you have to know about metadata:
What is а metadata? Metadata actually is “data about data” or in other words it keeps additional information about the source. When search engines first began appearing in the 90′s, they heavily relied on this meta data to figure out what pages were about. Nowadays metadata can be used for finding useful business information by making precised content drilldowns.
How we visualize metadata? In the process of building Sensika we implemented an easy way to access and to make drilldowns using metadata. Even a non-professional can dig deeper into content and investigate business cases using the extracted metadata. One of the the most common retrieved metadata includes information about language, location, keywords, media type, organizations, originator, people, publisher, sentiment score, talking about topics, target geo-location, venue and all entities related to the particular piece of content.
How a user can benefit from the metadata? The additional refinement of the returned results through metadata can bring useful market and segment insights. For example: you notice a trend towards increasing the volume of articles with negative sentiment. Metadata can give you the answer to the questions: who are the publishers, the authors and how influential they are. Also you can find information easier by making content drilldowns based on metadata such as location, gender, Klout score, organization and many others.
The bottom line: Once extracted and made easy accessible, it’s a matter of business needs how a user can benefit from metadata.