Sunday, January 26, 2014

Predict fires in the cities by using big data


The best-known examples of big data implementations is the analysis of customers' behaviour enabling to support the customer during a buying process very effectively.

So I was very happy to read about the usage of big data increasing the proactive measures in fight against the fire in the cities.

Based on certain factors like neighbourhood income, age of the building, electrical issues etc. authorities in New York City can identify the critical buildings with high probability of fire.

In the mean time the NY fire department has defined about 60 of these factors and based on them a ranking buildings database has been created in order of their risk of fire and which ones should be inspected first.
Before this big data analysis the NY firefighters inspected with high-priority buildings like schools and libraries more frequently. But all other inspections were random.

The new system makes it possible to reduce the number of fires and make fires less severe, according to the fire department.

Although one thing stays difficult...
How to evaluate the impact, the results and merits of the big data approach to prevent injuries or fire.

Sunday, January 19, 2014

EDI without Mapping



Our company offers B2B integration know-how that has been built up over 15 years - and deals with the entire process chain in EDI in different branches.

It does not matter in which part of industry, there is still the same challenge if two companies try to establish an electronic data interchange (EDI).

The most painful and costly step in the on-boarding process of a B2B integration are the mappings from one message to the message understandable by the enterprise applications (CRM, ERP, Business Intelligence, ...) of the particular business partner.

And it is not just one time operation (expense), every time the data structures in the enterprise has been changed the associated mappings have to be modified.

Another challenge is the dependency of the developed mappings to one translation tool, without any easy possibility to be ported into another translation tool. The mappings need to be rewritten completely new.

Wouldn’t it be possible to find a way to do B2B integration without having to manually develop any mappings?
Does it sounds like Sci-Fi to you? Not to me!

Just think about technologies behind such products like Siri, Shazam, Google Translator, OCR, IBM Watson, and many more.

In case of EDI we are talking about structured data based on identifier what the data fields in the source file map to the respective fields in a target structure.

One approach could be developing such an algorithm with enough sample data to test and optimize it until it reaches 99.999999% rate. Sort of "learning".

In my opinion the time is ripe and some products from different areas show us, what is actually possible today. So this type of transformation is certainly within reach!

... and the vendor that can master this challenge will have no competition. It will be possible to save a huge budget in B2B integration projects.

The one will be the winner!