Turning data into business value begins by capturing it. At first, we can use manual entry templates, but as the amount of data increases, we need automated data capture and recognition.
Automatic data capturing and recognition (or identification) means automatically identifying objects, collecting data from them and entering it to computer systems. All with out human involvement. Basic example would be an invoice. There are many ways to do this and our solutions are mainly focused on optical character recognition (OCR). It is suitable for printed text recognition.
Unstructured data challenge
There is structured data such as tax returns that have always completely same structure. This data is the easiest to capture and doesn’t need special applications.
However majority of data is in semi-structured (such as invoices, purchase orders) or unstructured (such as letters, contracts, articles) form. This data might have various different appearances. Or the data form can be completely flexible. Semi-structured and unstructured data needs special applications and solutions to be captured.
As the data capture process becomes established, we can implement data-mining solutions and practices. After that, we can consider using descriptive analytics–statistics that describe your collected data. At this stage we can implement neural networks, technologies such as IBM Big Insights and various artificial intelligence approaches.
We have delivered capture-phase solutions at the “Performed” and “Managed” capability levels (as per our diagram) for such organizations as the large Norwegian car service company, Bertel O. Steen, and the City of Lahti, Finland. Please take a moment to familiarize yourself with them.
Next read further how you can get more out of your data by enriching it.