SMEs might not include themselves in the Big Data and text analytics discussion because they think they are too small or don’t accumulate enough data to be considered ‘Big Data’.
But Big Data isn’t just concerned with volume, but also velocity (the speed of change and how quickly it needs to be processed to yield meaningful insights) and the variety (i.e. the various sources of your enterprise data including videos, images, social media posts).
So, if your company has volume, velocity and variety – Big Data should be important to you, and it will only ever become more so.
Text Analytics: What it Means
More than 80% of enterprise data is unstructured, meaning it does not follow a pre-defined data model and does not reside in a traditional column database. Examples include videos, business emails, audio files and social media feeds. Each of these data sources has such varying levels of data, it makes it impossible to assign consistent data mining principles to it.
Unstructured data also causes problems when trying to analyse it and transform it into useful and actionable information, due to its infinite variety.
Text analytics is the process by which companies can create meaningful outcomes from otherwise difficult to interpret unstructured data. This is done by;
- Introducing parameters for analysing unstructured text
- Identifying and extracting relevant information
- Transforming that information into useful, structured and meaningful data
How Text Analytics is Achieved
By extracting the when, who, where, what and why from unstructured data, companies can gain insights into who is talking about their brand and what they are saying and use this in a number of useful ways.
For example, an online retailer might discover that customers in a certain country buy more clothes in a particular colour that coincides with their annual national holiday – knowing this information enables the company to provide targeted offers and campaigns to optimise this spike in purchases.
SAP’s HANA analytics tools use two key components to interpret unstructured data;
- Linguistic analysis – segmenting (dividing the various inputs of text), stemming (identifying word stems) and tagging (word labelling)
- Extraction – identifying entities (names, dates, places etc.) and facts (relationships, events etc.)
The extracted data is then configured to show the real ‘voice of the customer’ and their sentiments about your brand as well as common mentions of things within your business environment. Combined, this data can show some powerful insights into your market that you may never have otherwise uncovered.
Text Analytics: Real World Examples
An online grocery delivery company with a 100% on-time guarantee. Due to the perishable items they are delivering around New York, they need to minimise the delivery times wherever possible.
By using SAP HANA, they are able to run at a 99% on-time delivery rate.
The biggest teaching hospital in Europe providing 150,000 inpatient and 600,000 outpatient treatments each year.
Charité now uses SAP HANA to analyse data from cancer and medical databases (including structured and unstructured data sources) to identify suitable candidates for life-saving clinical trials. They plan to add DNA to the analysis soon and will then be able to analyse up to 500,000 data points per patient.
Global animal nutrition leader, Provimi, has used SAP HANA to reduce their report processing time from 10 hours down to 2.4 seconds (including 2 years of data). Not only has this created more efficiencies for business users, but they can also make real-time purchasing decisions based on accurate data.
If you like what you see and want to leverage the benefits of text analytics for your business, let the ERP and HANA consultants at Blue Ocean Systems show you exactly how your current data could help you improve processes and drive profitability today.
Note: This story has also been adapted for publication in Steemit.