The PuppyIR repository is organised, as most SVN repositories are, into three folders: trunk, branches and tags. Each of these folders is detailed below with a picture of their structure and a short description of the key parts contained within them.
To checkout the whole repository (this is a large download of ~600MB) and browse to the top level of the repository use the following commands:
$ svn co https://puppyir.svn.sourceforge.net/svnroot/puppyir puppyir $ cd puppyir
N.B. the diagrams shown in this section are simplified, in that, except for a couple of exceptions, no files are shown; only folders. Also, standard Django application folders (like ‘site_media’ - for example) are not shown in order to make the diagrams easier to read.
This section is the main development area of PuppyIR, it contains the latest version of the framework and various applications (plus demonstrators) that make use of it. Following the diagram below, the key sections of trunk’s contents are summarised.
This folder contains the latest version of the framework, the test suite and the documentation (both the source and ‘compiled’ versions). The main sections of this part are summarised below in terms of their contents (see Service Architecture for a conceptual description from the perspective of a search service based architecture):
build and setup.py: are the build directory (for when installing the framework) and the Python script to install the framework.
test and unit.py: contains the test suite directory and the Python script for running the tests, please see: The PuppyIR Framework Test Suite for details of this component.
In the trunk there are two demonstrators which serve as showcases for the PuppyIR project; these demonstrators are described below.
This demonstrator, also known as the Emma Search service (EmSe), is being built for Emma Kinderziekenhuis (EKZ), which is part of the Amsterdam Medical Centre (AMC). At the EKZ, children have access to a dedicated information centre as well as a dedicated bedside terminal. A user study carried out by hospital staff from the information centre has uncovered that children are reluctant to engage with the physical information centre (depending instead upon a family member or carer) and so, EmSe is designed to make use of these bedside terminals to allow them to access this resource via the web.
The motivators behind this demonstrator are, therefore, to:
EmSe assists the children by providing appropriate query suggestions, simplifying difficult content and filtering unsuitable content based on age appropriateness.
The Museum Demonstrator creates an interactive museum visit using advanced technologies such as multitouch tables and marker tracking, creating the basis for additional data retrieval and filtering using the PuppyIR framework. Up to four users can use a multitouch table simultaneously to browse through the different exhibition subjects and together they determine the contents of an interactive quest.
Subsequently in a trail through the exhibitions users/players answer questions related to the chosen topics that have to be found. Throughout the museum various touch-screens equipped with scanners for reading and identifying the players are installed that when triggered present the questions and provide feedback to answers.
After all questions have been answered, the multitouch table provides further information about the visited exhibits.
You can view a video of this demonstrator in action by visiting: http://www.youtube.com/watch?v=b5zycfgqlKo
This folder contains prototypes made using the latest version of the framework. These prototypes are either completed or in the late stages of development and so are all in a demonstrable state.
These prototypes are explained in: Running Prototypes - please consult this page for more details.
This folder contains the University of Strathclyde’s experimental environment on collaborative search interfaces.
This folder contains standalone components and unfinished/work-in-progress prototypes.
AnSe this is an application that uses the PuppyIR framework to query, using the Bing and YouTube wrappers, and retrieve results in the JSON format. It is totally standalone as it contains its own, reduced, local copy of the PuppyIR framework.
conf demos (framework and hospital) these are early versions of a method to allow for easy configuration of these resources.
Interns: a application called ‘sniffer’ created by student interns working on PuppyIR, this application consists of: a search application similar to BaSe (see below for more on BaSe) and an automated logging application called ALF (Automated Logging Facility).
Student projects this contains applications made by students studying the Internet Technology module at the University of Glasgow. At present it only contains the original version of the aMuSe application (the new versions, as detailed earlier, can be found in trunk) using an old version of the PuppyIR framework.
There are several prototypes contained within the aforementioned ‘working’ folder. These prototypes provide further examples of how to use the framework but remain in-complete and as such, may contain flaws and/or not fully function.
N.B. Once completed, these prototypes will be moved to ‘trunk/prototypes’.