Exploration: Possible update to Map and Define Glossary in Author
When doing my PhD in computer science at the University of Southampton on academic knowledge work, one of my surveys showed strongly that a key issue for academics is to see and work with connections. That is exactly what this Map improvement is intending to do.
How can we better see connections? By making them clear. This may seem pithy or even self-evident but this is not always done in a clear way in academic, scientific or scholastic papers, as evidenced by my studies. Here I present an approach where the user defines glossary terms, and in so doing connects the terms, through the definitions, which can be seen mapped out in a dynamic Map view.
How this approach differs from others
It primarily differs from other in how we deal with mess. No images or icons and no persistent lines to make the screen messy. It also allows the user to cleanly and clearly define what they care about in a way which the reader will also benefit from.
In the current Map view the user can move defined glossary terms around to group them. When the user selects a term, any other terms which are referred to in its definition get a line pointing to them. This gives the user an opportunity to examine how the definitions are correct and can update them as new understanding emerges.
The defining of glossary terms is a means of making as clear as possible what one is trying to communicate, cutting through much ‘verbiage’. The associated steps include:
Define The purpose of defining glossary terms is initially for the author to clarify their thinking by exploring how the key concepts in their work are related. New is that all defined glossary terms will be visible at all times.
Publish The middle job is the part where the author prepares for publication–that is, making the document public for a reader–where the focus is on making the document digestible in an efficient and pleasant manner. New is the ability to annotate the Map with headings and explanatory text, as well as deciding what should be visible and what should be hidden.
Read The final purpose of the Map to help a reader understand the result of the author’s thought process. The job of the software ‘Author’ is to augment the authorship process and the job of ‘Reader’ is to help the reading process, hence this document does not cover reading too much, only how to prepare for it.
Currently the export on Publishing to PDF includes a Glossary, so that in itself should be useful.
Another benefit could be that the glossary terms can be treated as relevant keywords for search because they are demonstrably key for the author.
Letting the author define the terms they work with, with plain sentences and basic tagging, and exporting this in machine readable ways, can provide rich ‘reading’ interactions in multiple environments later, which will be a different challenge to address and I expect both traditional reading and rich VR/AR reading to be able to take advantage of the connected data created.
The big change from how the Map currently operates is to show all Defined Glossary Terms in the Map view at all times, as well as letting the user add more metadata to the Define Glossary dialog (which will not be visible on the Map but which will be accessible to VR etc.)
We will also need a change name in Author from ‘Defined Concepts’ to ‘Defined Glossary’, in the menu and in the cmd-D dialog (which will be labeled‘Define Glossary’). This is to make the ‘Glossary’ term used in both Author and Reader. The removal of ‘concept’ simplifies it all a bit. We keep ‘define’ and ‘defined’ since cmd-d is useful as a keyboard shortcut and the main action here is the defining, not the concept.
User can still add text which is not in a Defined Glossary term, and it will be added in italic, which is useful for explanations or headings which does not require a definition.
- Remove option to select text and ‘I’ to make italic and ‘b’ to make bold.
- On export only defined glossary items which have been used in the document will be listed in the Glossary Appendix.
- No heading for Definition, instead show grey text which instantly disappears on cursor entry.
- Dialog text to have slightly more space between lines. Ideally same font as body font, but that is only ideal.
- Add option to enter a date or data range, same as in Citation dialog. This should ideally be able to handle BCE.
- Add a pop-up for ‘Type’ with options for:
- General (default)
- +, where ‘+’ opens an option for custom type. When a type is added added, it will remain an option for new definitions for the future.
- Show all ‘Defined Glossary’ terms/nodes in the Map view at all times, as opposed to now, where the user manually has to add. In the future we may allow for some filtering, for now show all, so that we can learn what filtering might be needed.
- It will still be possible to add text straight to the Map, user can double-click on the map to type anything. The typed text becomes italic if it does not have a definition. Once there is a definition it becomes non-italicised.
- There is an auto-initial-layout if this is easy to implement and maybe
- Select item and = (equal key) to select all of the same type (as defined in the Define Glossary dialog).
- On selected text’ h’ to hide that node(s). ‘h’ again to see all. If user selects text which is connected (in either direction) to hidden nodes then ‘h’ will un-hide the hidden nodes which are connected. Simply selecting a node which is connected to hidden nodes will show them but faded, same as the ‘from’ connection line.
- Select and ‘delete’ will open the Defined Concepts dialog and ask if user wants to delete.
- Ability to save layouts which can be linked to from the body text in the same way internal links can point to headings and images. The bottom bar in the Map view, on the left, will have a pop-up for the user to save layout and name them.
- Bottom right will have an option to ‘Show All’ which will reveal any folded/hidden items.
- Better parsing for connection lines. Ideally we need basic parsing so that ‘AR’ will not draw a line to ‘ARC’.
- Selecting text and dropping it onto other text makes the text dropped onto bold and becomes a ‘heading’ for the dragged text.
- To fold and unfold (hide/unhide) items the user needs to double-click on the bold heading.
- When thus folded, any lines to hidden sub-item will go to the bold heading.
- When un-folded, all the ‘children’ are selected. This allows the user to move the whole set as one.
- To edit a heading, the user needs to select the heading and ‘enter’/‘return’, same as editing other nodes.
- There will be a simple way for the user to ‘Copy the Defined Glossary’ from one document to the next, as a command at the top of the Defined Glossary list, under ‘Define’. This allows the user to have more than one Glossary.
- Related to this update of the Map is the issue of placing the .liquid document somewhere online and have the PDF’s Visual-Meta refer to it, so that all the information in the .liquid document can be used, for example in VR. The .liquid document has readily available JSON metadata so this is more efficient than Visual-Meta, when available.
- Storage of 3d coordinates.
- How to constrain the Map to still fit the 2D version, which perhaps becomes an authoring issue by making the canvas more flexible, after interaction in VR.
- How to map different Glossary terms from different people/documents in a single VR space.
- How to specify what Glossary terms should be available even if the host document is not open, if any.
- At least as an image (for each saved layout).
- Better if it can be the size of two pages but compressed into one, turned 90 degrees sideways but has Visual-Meta to tell the reader to show it turned 90 degrees back.
- Best would be to also have the Map engine from Author available in Reader.
- Allow for easier access to see what words on a page are Glossary terms. In ‘Reader’ currently you need to select text and do ‘Find’ on the text to see the Glossary definition, or read the definition in the Glossary Appendix.
- Make available for augmented environment interactions.
- Allow readers access to similar Map view as authors have.
- Ability to pin or mark as core/high value in the Map view
Define Glossary dialog refinements
The Define Glossary dialog becomes taller, by 50%.
A ‘Private Notes’ section for any user comments which the user does not want to export.
https://js.cytoscape.org can help but if not, then we can simply have a list on the side of the screen, maybe the left hand side, and user does initial and later layout.
(If that text was previously italic, because it was not a glossary term, that doesn’t matter, it will still go bold).
Bridging 2D & VR
The location data will need to be able to cope with 3D location data from VR HMD for location of defined glossary node, with the initial variable of z dimension when edited in VR.
Implementation issues in augmented environments
On publishing the document the user can currently specify in the Export dialog whether to have a cover page, table of contents and so on. The Map view should be part of this stage, for the author to specify how it should appear to the reader:
Implementation issues in Reading software
Once published, to PDF or VR, there will be opportunities for interactions beyond what I can outline today, though a few points might be worth mentioning:
Define Dialog Mockups
Define Dialog. Hegland, 2022.
Define Dialog with Date Entry. Hegland, 2022.
Define Dialog with Date. Hegland, 2022.
Map with Bold headings and Italic comment, 'Visual-Meta' selected. Hegland, 2022.
Curated, ‘Saved’ Views
As discussed, it should be possible to curate the map for publication, so that in this example sections have been folded in to the headings:
Curated. Hegland, 2022.
Here the reader can instantly see what Mark Anderson is connected to in this environment:
One Curated Selection. Hegland, 2022.
Curated views could highlight specific aspect, for example 'Organisation' could open up this view where much has been hidden but all the people and companies are listed:
Organisation. Hegland, 2022.
The hope is that the decision to show all defined glossary terms at all times as nodes, while letting the user hide nodes and specify saved layouts, will allow the author to really think through the work in new ways and the reader explore with less effort and more insight.
The possibilities of a Map view in VR with these attributes (dates, type tags and layouts with headings and annotations) can be a powerful set of data for building VR interaction explorations for us to learn what interactions will be powerfully useful in VR for text-based work.