Skip to content

Home

Jump to bottom
alanepaull edited this page Aug 5, 2019 · 18 revisions

Jisc Intelligent Campus: Investigation of data and information requirements

This wiki describes the work that Cetis carried out up to the end of July 2019 to support Jisc's Intelligent Campus R&D project, researching the use of data to inform future campus developments.

Our investigations were based on a range of intelligent campus use cases developed by Jisc. An important context of the project was that requirements were based on Jisc's successful Learning Analytics Service, including its exising repository of student data, the Learning Data Hub.

The main output of our work was an intelligent campus data specification, which forms the main Github repository. Implementers and technical staff may wish to delve into the technical details, and we would welcome contributions to the Github repository. You can raise issues on the Github issue tracker.

Wiki Contents

The wiki contains:

Outline of Cetis' investigation of data and information requirements

The starting point for our work was Jisc's list of use cases, supplemented by interviews with interested staff in universities and at vendors of HE systems software. We examined each use case in order to gain a deeper understanding of ‘the intelligent campus’ concept and to identify and list data requirements. We developed a process to analyse the use cases formally. This involved creating for each use case a description of its core functionality, brief descriptions of examples from the use case, and specific measures suggested or used in the examples. This analysis is available in the Evaluation of the Jisc use cases section of the wiki.

To help to structure and understand the data and information requirements derived from the use cases, we developed an information model with a diagram to support it. This model linked elements in the vendor landscape (services and products) with Jisc analytics products built on the Learning Analytics infrastructure, with use cases and case studies, and with the technical specification products, including Methods, Measures and their Context.

Elements of the data requirements information included a comprehensive List of Measures encountered in the examination of use cases, which we brought together in a Measures Catalogue. These ranged from ‘Unstructured Feedback’ to ‘Environment Monitoring’ to ‘Location’ to ‘Energy Use’. In accordance with our information model, we analysed each measure for the methods used to obtain the data, the contexts in which measurement was carried out, the basis of any algorithms used, privacy implications and sources. This information provided the list of data requirements. During this work, we referred to a variety of existing data standards and data sources listed here.

We then prioritised these requirements, in order to focus on those that were most advanced in activity within the HE sector, evidenced by our interviews with staff, supplemented by our own and Jisc’s knowledge, and likely to be the most immediately tractable in terms of data acquisition. Looking at practical data collection issues at the start helped us to keep our feet on the ground in respect of implementation issues. The areas we agreed to focus on were environmental sensing and wayfinding, typified by the following use cases (links are to our evaluations, which also contain a link to the use cases): Building analytics, Intelligent spaces for learning, Finding your way and I think I know the way.

Mapping and wayfinding

We developed a simple service point classification based on Open Street Map concepts. This classification gives providers a common list of tags relevant to physical locations and service points in the HE sector that can be readily applied to Open Street Maps directly or to similar or linked implementations.

Sensors

We developed a data model for recording various types of environmental data from a wide variety of sensors. The model describes the structure, entities and properties to record Measurements taken by a Sensor in a specific Context. It allows both minimal data for ease of data flows, and enriched elements where more explicit or descriptive data is available. It includes a draft measurement taxonomy, providing key/value pairs for recording various types of measurement.

Locations

For wayfinding, sensing, use of building analytics and intelligent use of spaces, we developed a relational model of locations and their characteristics. This model extends the existing Learning Data Hub, Unified Data Definitions (UDD) to include concepts related to the Intelligent Campus, primarily in the physical space.

Mapping to the Learning Data Hub

Our work on the specifications enabled us to map Intelligent Campus data to that in the Learning Data Hub. As an illustration of a practical approach to linking from the Intelligent Campus space to the Learning Data Hub, we wrote a detailed narrative example that supports relevant use cases and shows the use of the information model. This example takes a provider with environmental sensors in its buildings and demonstrates how data can be collected and integrated with systems supporting teaching and learning to create an “intelligent campus dashboard” to show, review and analyse space utilisation in detail at module and event level to aid better decision-making.

The Cetis team

Clone this wiki locally