Data Science/Analytics use case advice

Hi guys,

Desperately need advice on the analytics use case. I have around 60 odd BI reports (cloud based data reports for employees across globe). I need to identify similar ones so that the reports can be merged and number of data models can be reduced.

Data:

Excel sheet with following columns:

Column 1: Report ID (60 reports)

Column 2: Sub category -( reports have been categorized into 4 sub-parts based on usage )

Column 3: Table names which from which data is being fetched Can be more than 1.

Column 4: Names of users who use the report ( not more than 10)

Column 5: Report Field names - SI, CI etc = final columns that we arrive at, after using function etc on the data

Based on these columns/Data for each report, is there a way I can find similar or merge-able reports to reduce the number of data models.

Somebody suggested clustering, but wasn’t sure about it.

So is there a Data Science way/method that i can apply here with good enough accuracy. Any advice would be a huge help.

Thanks & Regards

Any help would be great

This sounds like a data engineering use case.

I’d do it as mentioned below:
Build a vector matrix between Report Name & Data Fields that are being used to build the report.
Once you have the cross tab ready, you can look at the total count (horizontally) to see which are the reports sitting together (sort by total desc). you can then classify the reports that are using the same data fields and then work on the backend engineering to reduce the data models.

© Copyright 2013-2021 Analytics Vidhya