The data to collect

The type of activity data that needs to be collected depends on what the data is to be used for. Often however, there are limitations imposed by proprietary systems that do not have APIs or published database formats to enable data collection.
In the Activity Data Programme, the major uses for activity data fell into:
  • recommendations for library resources and other resources
  • learning analytics to provide recommendations as to students at risk, so as to improve their results at university
The following table summarises the data collection strategies employed:
data description
Recommendation system for research repositories
DSpace repositories
Viewing and download information
Improve the videoconferencing service
Log data from videoconferencing bridges
Using AG Toolkit and IACOM bridges
Management information and to identify "at risk" students
All the events from the Sakai event table
Demonstrate correlation between library usage and student success
Library turnstile system
Library management system
EZProxy service
Student record system
Library and student records data
See collecting the data for a detailed description of the data and how it was collected at De Montfort University and the University of Lincoln.
Recommender systems to enhance the student experience
EZProxy service
student record system
Ebsco Discovery Solution
Library and student records data
Manage interventions with students at risk
CMIS - timetable; attendance
Banner - personal; course; grades
Honeywell - security gates
User - notes
Help Desk - Search; incidents
Xstream Portal/VLE - usage; assignment status /mark
Symphony - search; loans; fines
Resource Manager - AV loans; fines
Gmail - email, apps
Agresso - fees payment
Support humanities research by surfacing underused "long tail" library materials through recommendations
Talis Alto
Circulation data
Integration of activity data spread across various systems in an organization, and exploring how this integration can both benefit users and improve transparency in an organization
Diverse web servers
Server logs
Publication of anonymised OpenURL data for third party us
OpenURL Router
OpenURL requests
General limitations on collected data:
  • Data quality: Some of the data may be corrupted.
  • Data completeness: Some of the data may be missing.
  • In some cases data volume may be large (eg EVAD had very large log files generated by Sakai)