Dark Data Analysis

Dark Data is a kind of unstructured data. What's special about dark data is the fact that no one has looked at it previously as data. So special skills are required to detect such dark data in your organization and put it to effective use. It is also very difficult to convert it into structured data.

Cere Labs has built an AI platform called Cerescope. This platform is capable of converting dark data into structured data.

Structured v/s unstructured data

Structured data has a well defined schema, for example, a spreadsheet or a database table. This data can be used for analytics, prediction, automation and many other purposes.

Unstructured data, on the other hand, is not easy to use. It does not have a well defined schema. Email is an example of unstructured data. It contains text that is valuable, but no algorithm will directly work on it.

Types and sources of Dark Data

There is a lot of useful 'Dark Data' in your organization. Examples - pictures, CCTV videos, text documents, voice and so on. We can make sense out of all this and provide you with invaluable insights.

Video
(CCTV Feeds, Video recordings)

Images
(Mobile snaps, medical images)

Speech
(Call recordings, dictations)

Text
(Email, Documents)

Handwriting
(Applications, challans)

Waveforms
(Machines, sounds, medical devices)

Example Use Cases

Video Analytics ( People Counting )

Video Analytics means using basic video feed to produce insights about the business. Using AI technology, videos from ordinary CCTV cameras are analyzed for counting and tracking people. The actual metrics produced depend upon the business being analyzed. It can be used by a variety of businesses such as retail, security, healthcare and so on.

Image Analytics ( Stock Taking )

Stock taking is a painful and lengthy process. Now Dark Data Analytics can be used to automate this process. Images taken in the store can be used for counting and determining the actual stock.

Text Processing ( Lengthy Documents )

Those who have to read lengthy and heavy documents as part of their job know how troublesome that can be. Large amounts of text is a kind of dark data. Cerescope has capabilities that can help the users to quickly find answers or compare two documents.

Voice Analytics

There is a wealth of information in recorded speech such as customer center calls. This dark data can be used for customer sentiment analysis, to detect trends and measure response to your products.