poloclub / argo-scholar Goto Github PK
View Code? Open in Web Editor NEWLiterature Review Made Easy with Visualization
Home Page: https://poloclub.github.io/argo-scholar/
License: MIT License
Literature Review Made Easy with Visualization
Home Page: https://poloclub.github.io/argo-scholar/
License: MIT License
right-click empty space to deselect any nodes that may be have been selected. Then left-click a node, and then right click; the menu that shows up will not have “Argo Scholar” menu items
Symptom: when pressing left arrow key (or other arrow key)) on keyboard in search box, the graph also pans left.
Possible reason: the graph “view”currently supports panning by arrow keys.
Possible fix: the simplest fix could be to have the keyboard presses (KeyPressed events?) be “consumed” by the search box (thus will no longer propagate to graph view)
How is “Show 5 neighbors with highest pagerank” different or related to “Add 5 Paper Citations/References”? E.g., is there a dependency (e.g., need to first add 5 citations/references before users can get 5 neighbors with highest pagerank?)?
ArgoScholar is a really nice tool to explore the personal network for an own paper.
However, often one want to leave some personal notes to a specific citation (after reading it, for example to leave some comments how exactly it fits to your own paper).
It would be really nice, if those notes can be added as part of the window at the right side with the details of the citation.
Maybe it could be even more generic and one can add arbitrary key, value pairs. With this, for example a bibtex key can be added to connect ArgoScholar with preexisting bibtex files etc. or adds the ability to integrate ArgoScholar in some preexisting tooling for citation searches.
Problem : only add 10 citations based on relevance, 10 citations based on importance, etc, due to the page and pagesize parameters of the api call being fixed to 1 and 10.
Possible Solution: have multiple fetches and store the results in a promise
Currently, edge arrows are quite small and are hard to see.
Also, since arrows pointing both directions are both placed in the middle of edges, it's not easy for users to quickly identify an edge's direction.
Unpinning node doesn't work when any node is clicked immediately after clicking the unpin button. The background has to be clicked for the node to be unpinned.
Create a "New..." option under "Graph" which acts as "Clear all papers." Also create a popup that informs user of this action.
Improve usability and readability of color legend (we thought about doing this for a long time, but didn’t get to do it) to something similar to (doesn’t need to be identical of course) http://friendlycities.gatech.edu/data-interactive/ . If there are existing libraries we can just, there’s no need to reinvent; just use it
currently, right-clicking empty space (that’s not a node) causes a pop up menu to show. The popup menu should * not * shows because it should only show when at least 1 node is selected
Currently, the link opens in the same window of argo scholar, potentially overwriting user's current work/visualization
To reproduce:
For each paper node, also retrieve and save its CorpusID (as property and value).
This list is intended to be an archive of some UI issues on the Argo Scholar web app. Updates and fixes will be added regularly.
Dark mode color scheme is inconsistent in certain dialogs. (Dark mode disabled)
Selecting text is disabled in certain input boxes, e.g. the one in RenameSnapshot. (Might be a bug of blueprint.js)
Add graph by csv/gexf still exists in the graph menu even when they are no longer supported by Argo Scholar. (Solved in 39164a3)
The "done" button is still clickable when it has the CSS property of "disabled". (Solved in 2931054)
The automatic file download function is not implemented but the text states that
If your browser doesn't start downloading the file, you can manually copy the content below and save to a plain text file.
Graph Options card's width changes when scroll bar appears. (Solved in a2333e3)
Add a search bar that opens the current add paper menu
For paper nodes with 10000 or more citations, the citationCount is displayed as 9999.
Improve usability and readability of node size legend (that’s another thing we didn’t get to improve), so something similar to (no need to be identical): http://friendlycities.gatech.edu/data-interactive/
Are there ways to add multiple corpus ID at the same time?
The tool prompts incorrect Corpus ID sometimes even when the ID is correct.
To recreate this:
How to add more than 5 citations or references for a paper node in Argo Scholar? And why limit to 5?
In Argo Scholar, we need a way for users to “link back” to Semantic Scholar (from a paper node in Argo Scholar).
Polo: Oh, I just saw that googling “argo scholar” would show the result of tool/site with little description — see attached images from Google and Bing. Could you help improve it so the description is more informative, and importantly show up in search results (e.g., by taking hints from Google Scholar, Semantic Scholar, Connected Papers, etc.)?
To follow the convention of popular applications, reorder the menu items to:
Save Snapshot
Share Snapshot
Steps to reproduce:
create empty graph, then add the paper node with corpusID 4891043 (or other corpusID)
right click paper node and click "add 5 paper citations" (and the error message pop up once for each citation)
When zooming in and out, animate the transition, to remove sudden movements (which disrupts people’s visual context). Relevant examples (found via googling “d3 zoom animate”):
https://observablehq.com/@d3/programmatic-zoom
https://observablehq.com/@d3/smooth-zooming
ArgoScholar looks extremely promising to manage paper citations. However, for this, I would like to keep the citation graph as part of the overall paper repository, so I need to export it.
Currently, when pressing "Save Snapshot" a hint comes with "If your browser doesn't start downloading the file, you can manually copy the content below and save to a plain text file.". While the copying works, it requires more work than a plain download would cost. The hint suggests, that at least some browsers open a download dialog but I did not find any (I tried Firefox, Chromium, Qutebrowser).
Does this open a download dialog somewhere?
Also, do you know of some possibility to make the whole opening/edit/saving process more comfortable? Ideally, I would like to have the citation file laying together with the remaining files directly in the paper repository for which I need the citations. Ideally, I would open the file, edit the file visually and "just safe it" (or even have an autosave feature). Even with the possibility to download the file directly, the browser directly is not really helpful in opening and saving local files.
When saving the graph as a persistent link on the server. Is there a possibility to get the raw string easily with a script?
Sharing an Argo Scholar snapshot should create an URL with the prefix that starts with https://poloclub.github.io/argo-scholar/...
Currently, when adding 5 citations/references for a paper node, those 5 (new) papers are determined by the order (unclear by what measures) determined by semantic scholar's API.
A better way to add the 5 papers would be to first sort all the citations/references by relevance (determined by SS), citation count, etc. Then, add the first 5 papers (that's not already shown in the visualization).
Currently, when user mouse-over a node, all opacity of other nodes (that aren't mouse-overed) is reduced, allowing the edges behind them to show throw and can look visually overwhelming.
Possible solution is to increase color lightness (e.g., by first getting the HSL color space representation of a node's color, and then increasing the "L" value). See attached image for how example lighter node colors.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.