Карта сайта
Версия для печати

Управление социальными мастер-данными и будущее бизнес-аналитики

7 марта 2013 На сегодняшний день управление мастер-данными (Master Data Management, MDM) является проблемой для многих компаний. Тем не менее, в результате стремительного роста популярности социальных сетей среди клиентов, компаниям приходится иметь дело с другой проблемой (и в то же время возможностью) в работе с Большими Данными в реальном времени. (Материал опубликован на английском языке)
In traditional MDM we aim to optimize the identification and descriptions of the who, what and where in traditional systems of record. Basically we handle our own products, our present suppliers, our current customers and known prospects and the related locations. When moving on to Social MDM we aim to link those entities to the who, what and where in systems of engagement so we may better handle descriptions of our own products, collaborate with suppliers and follow our customers and known prospects footprint in the digital world.

That may lead to the conclusion that what we can’t relate to our own products, our present suppliers and current customers and known prospects may be categorized as the famous noise in big data. But that may be a big mistake. In the future much competitive intelligence work will be data mining in social data. Social MDM may very well prove to be a fantastic new weapon in the social intelligence part of competitive intelligence by facilitating (near) real-time analysis and fast decision making. However, in doing that we need to embrace more entities than we most often cover in current MDM implementations.

Besides keeping track of our own products we need to keep track of our competitor’s products and relate his products to our own product hierarchy management. We also need to reflect social graphs and the result of network analysis back to party master data hierarchies. Thereby we may have an ongoing loop of exploring social data, analyzing social data and making decisions.

Doing that in a cost effective way requires a close relationship between governing our internal master data, utilizing traditional external reference data as product classification systems and available location and party reference data and the emerging capabilities of handling digital identity resolution in master data management solutions.

Source:  informatica.com