Scoring model/formula to rank based on accuracy of data sources

#1
Let us say i have 5 different data sources. My aim is arrive a score to arrive the accuracy of this data sources.

It is like below
1. Refer the section Data confidence calculation in http://ibmdatamag.com/2014/07/measuring-confidence-in-data/
2. https://www14.software.ibm.com/weba...t=swg&cmp=ibmsocial&cm=h&cr=crossbrand&ccy=us

My current approach is to prompt the user to answer the below attributes,

What is the frequency addition of new records?
Will there any human intervention in storing the data>
Will it get updated if there is any change of event?
And some more like this....
Then i want to use some model/formula to rank the most accurate data source. Please suggest if any.

So after that i will integrate all the sources of data. Assume there is some unique id exist to join the different data sources. So if a unique id ABC is present in multiple high accurate (which we calculated in above step) data sources i will give higher rank to that id rather than a id YYY present only in low accurate data source. For this which ranking model can be used?

Kindly provide your suggestions of the ranking model to calculate data accuracy and then use this score to rank the common attribute of integrated data. Do we need to use two different approaches/models for this.

thanks.