In Entity Group’s book “Crossing the Data Delta” we define the purpose of MDM as “to facilitate the generation of a single version of the truth for your organisation’s most important data” we also say that MDM “could be viewed as the engine that powers high-value solutions”. MDM is a technology but the reasons for doing master data management in your organisation should always have the delivery of business value behind them.
The linking of technology to business value means that a successful MDM project must gain broad business stakeholder acceptance that a particular configuration of a probabilistic matching engine is delivering accurate and trustworthy results. However, this can be challenging because algorithmic matching is often a subjective issue. One way to address this challenge is to involve the business users during the implementation phases of a project – for example in tuning the algorithms via “sample pairs” exercises. The process of collaboration helps to inform algorithm development and build the users’ confidence in the algorithm output.
• Tuning the IBM probabilistic matching engine (PME) algorithms that can be used to reach a
consensus within the business community
• Reaching and agreeing acceptance of the final algorithm configuration
Please fill out your details below to download the technical brief and sign up for future updates: