Yes. It can be.

Yes it can be. But most important fact is determining the criteria. I mean determining the important factors which are playing crucial character on institutional theory. These things identification in necessary.

Data mining can identify patterns but what will actually be the large amount of data, it is questionable cause institutional analysis varies. It is not like particular machine or singular event or collection of same type of events. That is the challenge I think.

Yes. It can be.

I am highly certain it can be integrated into institutional analysis.

The concept of "big data" should not be so scary because in the long run once you are able to apply simple data analytics techniques to a dataset of any size (large or small), the algorithms are transferrable and usable in the context of big data.

With regards to your question also, take a look at these resources:

• Sun, H. (2008). Implementation of a classification algorithm for institutional analysis (Doctoral dissertation, Lethbridge, Alta.: University of Lethbridge, Faculty of Arts and Science, 2008).
• Yu, C. H. (2013). Data Mining as an Extension of EDA and Resampling.
• Ahonen, P. (2015). Institutionalizing Big Data methods in social and political research. Big Data & Society, 2(2), 2053951715591224.

This presentation also, gives a clear picture (click next page to move to the next slide):

Big Data in the Institutionalization of Social and Political Research

• https://blogs.helsinki.fi/computational-politics/2015/03/12/big-data-in-the-institutionalization-of-social-and-political-research/

yes , Data mining can revealed interesting patterns like academic performance etc.

Yes... this is possible. you can take a looks at one of the 5V of Big Data which is "VALUE", which has to do with mining for data and analyzing it. the result is getting patterns and trends which can be used by institutions.

Big Data analytics with Predictive models can improve institutions

Yes,it has become part and parcel of Institutions as that generate huge data and try to extract patter out of the data in order to compete their peers.

Thank you for the insightful comments. I will take it into account in my next project. Best, Leandro.

big data becomes a trend for institutional analysis if the data does not have many differences or many directions in the same time series.

if the data has many directions in the same time series, we must study only the last part has the same direction in the time series. we can determine this part by using RMSE ( root mean of standard error ).