User Suggestions and Comments
Dear Knowen users:
We are looking for the ways to improve Knowen, and will be grateful for your suggestions. Please add them here, no matter how minoror ambitious, to create a record and possibly have a discussion about them. A suggested format:
Date: july 22, 2019
I cannot create any new node at (-2) level node at https://knowen.org/nodes/24392?path=1-24392
Please help - fix or suggest a reason for the problem.
I only hack this problem by creating a node at (-1 level) and then changing, which is very inconvenient.
This is kind of a typical problem. Due to some unclear reasons and only in some parts of the tree, nodes are impossible to create.
The error message is not insightful or helpful.
Date: sept 30 2018
a pretty annoying bug or non-report -"Please, correct the following errors: Parent nodes is invalid"
Not clear what to correct!
Date: may 25 2018
Take a look on the connections between Knowen and Knowledge Web of James Burke
Burke is developing a mobile app called Connections which aims to allow users to make searches in Wikipedia in ways that could lead them to develop mind associations and connections of apparently unrelated fields of knowledge or topics. Surprises, anomalies, and unexpected perspectives on a search can emerge from using the app. According to Burke, the Connections app is an alternative and innovative method to more linear internet search engines such as Google.
Date: may 25 2018
the left column "Dashboard and related pages" should be open by default ( for new users to find their orientation quickly!)
for the same reason, "graph of pages" should be open in a mini-format by default (possibly below the page structure (left column))
Date: may 25 2018
By: Mihail Turlakov
- it should be possible to subscribe to any part of sub-tree
- is it possible to refer and link to sub-section of the page, like references to sections in Latex?
- harder task: good interactions between questions/answers with the main page?
a thought: can you integrate with Wikipedia somehow? In particular, is it possible to build DAG-tree by applying PageRank to Wikipedia? Clearly, the links from lower ranks back to higher ranks will need to be dropped. In other words, Wikipedia can have two-way links and arbitrary topology. Some kind of filtering and layering will need to be applied for Wikipedia-Knowen transformation. What happens with the information content? Encyclopedic knowledge (Wikipedia) versus Scientific/Systematic knowledge (Knowen) - quantification, social impact, scalability,... and other consequences and comparisons?
Answer (Ivar M): There are definitely attempts to reveal implicit hierarchies of Wikipedia, e.g. Scalable models for computing hierarchies in information networks, Baoxu Shi, Tim Weninger. Still, Wikipedia was (and is being) developed as an ad-hoc network, so it is not always natural to try to force it into a uniform hierarchy. But it could be nevertheless a good first approximation, which with the right incentives could evolve into a well-balanced hierarchy à la knowen.
Knowen is a good tool for systemitizing your own knowledge. If Knowen-tree of global knowledge is not quite unique, what can Knowen be beyond the tool?
Answer (Ivar M): Hierarchy does not have to be unique. The same topic covered for a mathematician, physicist, doctor, or engineer can be structured somewhat differently. Just like textbooks can be structured differently. However, in a dynamic environment the hierarchy(ies) can evolve to self-fulfillingly satisfy the particular community needs (be it large or small community). The smallest possibly community is a collaboration, of 1-2-3 participants. There need not be a conceptual chasm between closed collaborative group and a broad community, even though the incentives for participation may be different.
Date: 15th Feb '18
By: Phil Crowley
Idea: Automatic highlighting of bracket pairings in tex would be amazing. It is hard currently to debug missed delimiters.
Efficient Graph Search:
As Knowen grows, finding a certain node or topic becomes harder. Currently, if one wants to attach a node to a proper parent node, one has to use a visual interface to the whole Knowen graph. It is really inefficient, especially when the graph grows further.
– Comments should allow formatted text (as in nodes), with full LaTeX/MathJax support.
– It should be possible to reply to, or 'mention' a specific Comment or person, by referring to a Comment's unique ID (e.g.,
@172736639) or a person's username (e.g.,
@JaneDoe). Hence, each comment should have a referable unique ID (perhaps, somehow obtained from the node ID).
– Comments seem to be unchangeable once inserted; it is also not possible to remove a Comment. These possibilities should be provided (as in stackexchange). I suggest to allow editing/removing comments for a certain time period, like 30 minutes, or so.
– To organise scientific discussions, Comments should have a tree-like structure, like Reddit. Then it becomes much clearer to whom one is replying, or from whom one is asking.
It would be nice to embed the "Discourse" system into Knowen for improving the comments and discussions. The source code is open and can be found here: < https://github.com/discourse/discourse >. I believe it would bring a 'stackexchange' experience to Knowen, with all its tried capabilities.
Currently, the Knowen's Comment section is very rudimentary and insufficient. I would like to see a possibility for multiple (but separated) discussion “threads” related to a certain node. The discussions should have a tree-like structure, with image-embedding, reStructuredText and MathJax enabled.
It would be also very nice, imo, if all the comments (on all nodes) can be seen in a separate dedicated Q&A page, to be accessed globally even without directly visiting the corresponding nodes where they were inserted. It will help quite a lot to find relevant comments to one's topic of interest, and thereupon, relevant nodes.
It should possible to control the details of how an image is displayed on a node; e.g., the width & height and the caption of the image. For instance, via
<img src="https://imagebase.com/12345.jpg" width="200" height="100" />.