Request for information

Some interest has been expressed in the development of an administrative tool to draw together and present usage statistics for UC Davis Sakai/SmartSite. Before moving forward, we want to get people's opinions on what they would like to see happen in this direction.

Primary interest groups/views/consumers of Sakai usage statistics

There are several different groups that may have both overlapping and non-overlapping requirements for assessing the actual usage of Sakai on a campus-wide basis.

Technical (developers, dbas)
- to predict growth, evaluate long-term resource needs; to quantify current demands and identify pressure points on the system as a whole
Communications/Marketing
- to track adoption and quantify user interest in the various components
Management
- to determine strategy based on current usage; to target specific groups for adoption

Types of data analysis

Potential solutions will probably fall into at least one of these two areas.

Real time event monitoring
Data-mining of existing usage information in the database

Most likely, we will encounter some constraints when it comes to mining the database on account of the sheer quantity of data involved. Tables that contain event data are frequently truncated, so any long-term monitoring would require some aggregation and archiving of data. Also, depending on the complexity, it may be necessary to batch some of the queries and/or denormalize some of the assessment data.

What's already out there?

A group in Portugal has put together a tool called SiteStats, which is currently marked as a Sakai Contrib. It provides what is effectively an instructor-level view of usage on a site-by-site basis, which does not at the moment seem to fit exactly with what's needed here at UC Davis. But it may be possible to make use of some of their code or user interface.

– Confluence Docs: http://issues.sakaiproject.org/confluence/x/Gz0
– SVN for v2.1.x: https://source.sakaiproject.org/contrib/ufp/sakai_2-1-x/sitestats/

Indiana has a tool-by-tool break down of their usage by semester including percentage change in usage since the previous year. Not clear on whether this is a one-time data mining action or a tool they have implemented.

UCD Functionality Wish List:

At this point in time, the following is just a wish list because we don't know yet if it's possible to determine all these statistical data points from the currently collected data in Sakai.

How many unique users have access sakai
How many course sites
How many project sites
Hourly concurrent users peaks
Most accessed course sites (top 10)
Most accessed project sites (top 10)
Most used tools in course sites (either by tool presents in sites or usage frequency)
Most used tools in project sites (either by tool presents in sites or usage frequency)
etc.

Once we have a complete UCD functionality wish list, we need to determine/research, which functionalities can actually be implemented. Feel free to add any comments below.

Browser not supported