Today, data mining is mainly connected with achieving a business benefit through the advanced analysis of structured data.
In the banking industry, for example, you take data such as "average account balance", "account balance-turnover ratio", "gender", "locality-district" and "age", applying complex statistical and non-statistical methods in correlation to records of sales, and the result is information such as "probability of a positive reaction to an offer of an X-product through the Y-channel", or "probability that they will leave in 6 months" for each customer.
A few years ago, non-structured data such as texts, voice records, website or tweet content, image information, etc. also became an object of advanced analysis. This type of analysis is usually called, e.g. "text mining".
At present, the importance of analytical procedures combining structured and non-structured data is increasing significantly. The business benefits of data mining is seen, for example, in increased income from higher sales, lower customer loss ratio, detection of fraudulent behaviour, reduced financial or reputation risk, etc.
In the commercial area, data mining brings regularly measurable financial revenue – cost savings, increased sales or minimised risks. It is often possible to visualise the financial effects of data mining – for example, the cost savings are related to the fact you can directly contact potential customers with an increased probability of positive response, while those with low probability will be left aside.
Sometimes the financial benefit as such is not the goal of data mining, but data mining is used for "deep insight" to produce source information for strategic decisions. For example, text mining has been used to select eight or ten key typologies from several thousand text responses from a survey which asked respondents "What should be improved in our company".
KOMIX s. r. o. connects the areas of data mining and web/text mining. The team responsible for the solution has extensive experience in using data mining for real achievement of business benefits. We do not search for technical solutions but for the benefit/cost ratio which will be optimal for the customer. From the view of technology, we offer implementation with use of state-of-the-art commercial tools (IBM SPSS, SAS, Autonomy IDOL), as well as with use of open source tools.