2.4 Naming Convention While segmenting the data, all numerically named slides retain their filenames, e.g. 1.jpg, 2.jpg. The title slide named title.jpg should be renamed to 0.jpg. All PowerPoint slides named hoy.ppt.slideX.jpg should be renamed as follows: Depending on what location they have been taken from, the slide should be named after the last numerical slide with the addition of a sub-section number. For example, in the sequence {67.jpg 68.jpg hoy.ppt.slide6.jpg hoy.ppt.slide7.jpg 69.jpg}, hoy.ppt.slide6.jpg would be renamed to 68-1.jpg, and hoy.ppt.slide7.jpg would be renamed to 68-2.jpg, so that the final sequence reads {67.jpg 68.jpg 68-1.jpg 68-2.jpg 69.jpg}.
2.5 Directory structure While segmenting the data, slides are moved from their original directory into a new directory tree, which can be read by CVNView. Following the first- and second-level classification from above, the correct directory structure should follow the outline to the right. The top-level directory must not carry any specific name; in the case of the example data, we have used w4111-8 to denote course w4111, section 8, and w4118-8. For any media type that has not been used in a given class, the directory must not be present under the top-level directory.	Figure 3: Class Data Directory Structure
2.6 Sections: Breaks in the Sequence of a Topic It is not unusual for a sequence of slides belonging to a given topic to be non-continuous. This would occur in a scenario where the instructor is, for example, working on one board, moving to some other board to explain some side note, or moving to a PowerPoint slide, and then moving back to the original board to continue the discussion. In this case, the slides should still be strictly separated by context and/or location. For the above example, the slides that the instructor is working on first are placed in directory "1", the side-note slides are placed in some other directory, and the final slides are again placed into directory "1". CVNView has been designed to read the contents of a given topic, and then, using the numerically enumerated slides, it will generate sections, where each section encompasses one consecutively enumerated sequence. For example, if the contents for a topic were to be {10.jpg 11.jpg 12.jpg 70.jpg 71.jpg}, then CVNView would generate 2 sections within this Topic: {10.jpg 11.jpg 12.jpg} and {70.jpg 71.jpg}. The reason for this method of dividing up slides is two-fold: Firstly, in order to preserve the order of slides over time, we need to cleanly break the sequence of all slides, even if they belong to the same topic. The graphical user interface has, after all, been designed to stitch together the pieces, so as to provide a meaningful display. Secondly, following the idea of the Forum Romanum Method, students may remember that one board was used extensively as some reference, meaning that the instructor would frequently break a topic to work on the reference board. This behavior is visualized in the graphical user interface by means of the sectioning of topics.