Skip to Content

Learning Analytics, how to implement

I'm in the Business Intelligence Department (BI 4.2, all WebI), but practically speaking we're the Reporting Department, your basic operational type of reports - lists, lists, and more lists.

I'm at an institution of higher learning, and a new project has come up for the Math Department. They want to know relationships between courses, grades, etc.

Examples:

- if someone gets a D in Calc I, what's the likelihood of graduation? with various permutations, like taking Calc I again
- what's the likelihood of someone getting a D in Calc I, getting a D or F in Calc II
- for placing incoming students in Pre-Calc or Calc I, what are the factors that indicate success? such as Verbal SAT

So I think I've targeted the right discipline (Analytics), but not sure where to take this projec

Any thoughts on where I could take this with simply WebI? Or maybe Crystal or some other BI tool?.

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

5 Answers

  • Best Answer
    Jun 07, 2017 at 07:05 AM

    My point of view is the following:

    With SAP Lumira or SAP Analytics Cloud, you have ways to forecast the evolution of a value in the future, based on historical data. These are simple trends that can be added on top of the visualisation of your historical data and this is powered by SAP Predictive Analytics behind the scenes. This is not provided out of the box by Webi.

    However, to make accurate forecasts, you typically need to take into account not only the date & the evolution of the value you are looking for but also any external conditions that can influence the value you want to predict.

    To take a concrete example, predictions of ice-cream selling typically depends on the forecasted weather conditions :-).

    So my advice would be to have a look and evaluate what you can do in BI. But if you want to go for the *real* thing, you need a full-fledge Predictive Analytics implementation. We provide a 30-day trial of SAP Predictive Analytics which can help you evaluate the capabilities of the solutions.

    Add comment
    10|10000 characters needed characters exceeded

  • Jun 09, 2017 at 04:13 PM

    You are welcome !

    BI = Webi, Lumira.

    BI + Predictive (for the business user, this is smart data discovery) + Planning = SAP Analytics Cloud

    Predictive (for the citizen data scientist or data scientist) = SAP Predictive Analytics.

    Kind regards

    Antoine

    Add comment
    10|10000 characters needed characters exceeded

  • Jun 06, 2017 at 02:14 PM

    So I was able to answer a question with WebI:

    "What is the Graduation rate for students receiving a D in XXX courses, where the student did not repeat the course?"

    pretty simple, actually, but did require some thought.

    As per normal, the key is to ask the right question, then search for the answer.

    So I can obviously answer some questions with WebI, but just wondered what do you think about the generic question, i.e., predicting outcomes by looking at history?

    For this, would you hire a consultant and build a Predictive Analytics system? Do this in Lumira (which I know nothing about) or something else? or brute force this in WebI?

    Add comment
    10|10000 characters needed characters exceeded

  • Jun 07, 2017 at 01:09 PM

    Antoine,

    very good analysis, thank you.

    I assume when you say "evaluate what you can do in BI" you're referring to WebI, Lumira. And is "SAP Analytics Cloud" and "SAP Predictive Analytics" one and the same?

    Add comment
    10|10000 characters needed characters exceeded

  • Jun 08, 2017 at 02:15 PM

    Best as I understand it, there are two SAP related Products

    - SAP Cloud Analytics - the "basics" of Analytics (except Analytics is an advanced topic)

    - SAP Predictive Analytics - industrial strength Analytics, typically "Big Data"

    Add comment
    10|10000 characters needed characters exceeded