Improve the Graph View
Sankari Nair
There is a lot to do to improve the utility and UI/UX of the Recall knowledge graph. Please add suggestions on how to improve it below.
Log In
R
Ritaban Biswas
the UI/UX could be similar to obsedian where you can colour code your nodes and adjust spacing with the fluid positions of nodes.
M
Matthew Gries
What InfraNodus is doing...do that!
Sankari Nair
Merged in a post:
Explanation of / Control over: the appearance of links in the Graph view.
M
Matthew Mensley
My graph view is a large web of colored lines linking actual pages I've summarized, plus autogenerated cards. Unless I'm missing something, there's no purpose to the colors - so I'd like either:
A - an explanation of what the colors signify if there already is a color coding system in place (I didn't see any mentioned in the onboarding)
B - control over the colors used, to better separate the types of links between cards (e.g. direct link, topical link..)
Sankari Nair
Merged in a post:
Make the recall graph way more powerful like Gold Standard Infranodus.ai
s
schroell
Graph Functionality Comparison: Recall vs. InfraNodus - Compact Analysis
## Current Graph Navigation Problems in Recall
Recall's graph visualization is currently
practically unusable
for knowledge exploration. The interface suffers from fundamental usability issues that make it ineffective for understanding conceptual relationships. Users report that even with modest node counts, the graph becomes an unnavigable "rats nest" that reverts to a messy default state whenever you navigate away.### Core Limitations in Recall
- Manual Connection Dependency: Relies entirely on existing forward/backward links rather than analyzing content relationships autonomously
- Poor Visual Navigation: Graph interface lacks intuitive exploration capabilities and stable positioning
- Limited Automatic Analysis: Cannot independently identify conceptual relationships from raw text input
- Unstable Layout: Graph positions reset when navigating away, making exploration extremely frustrating
## InfraNodus: Superior Architecture
InfraNodus
() demonstrates the gold standard for knowledge graph functionality that Recall should aspire to achieve:### Autonomous Relationship Detection
InfraNodus accepts
any text input
- documents, web pages, or collections of content - and automatically analyzes co-occurrence patterns to create meaningful knowledge graphs without requiring pre-existing links.### Advanced Graph Analytics
- Network Science Foundation: Uses centrality measures, clustering algorithms, and community detection to identify influential concepts
- Content Gap Discovery: Identifies structural gaps between topics that represent unexplored conceptual bridges
- Multi-Scale Exploration: Seamless zooming from overview to detailed relationship examination
### Superior Navigation Interface
- Stable Positioning: Graph maintains user-arranged layouts and configurations
- Interactive Filtering: Real-time adjustment of graph structure based on various parameters
- Contextual Insights: AI-driven suggestions based on structural analysis rather than basic text matching
## Development Recommendations for Recall
### 1. Implement Autonomous Content Analysis
Priority: Critical
- Develop text co-occurrence analysis similar to InfraNodus that can process any content input and automatically identify conceptual relationships
- Move beyond dependency on manual linking to create intelligent, content-driven connections
- Enable bulk processing of existing Recall content to retroactively generate relationship maps
### 2. Overhaul Graph Interface
Priority: High
- Fix fundamental navigation problems - graphs must maintain user positioning and not reset when navigating away
- Implement intuitive zoom and pan functionality with stable node positioning
- Add filtering capabilities that actually enhance rather than complicate exploration
### 3. Add Graph Intelligence Features
Priority: Medium
- Implement centrality measures to highlight most influential concepts
- Develop gap analysis to show where knowledge connections could be strengthened
- Create AI-powered recommendations for content that would fill structural gaps in knowledge
## The Killer Feature Vision
The target architecture should enable users to:
- Input any content(text, documents, web pages) into Recall
- Automatically generatemeaningful relationship maps without manual linking
- Navigate intuitivelythrough stable, persistent graph layouts
- Discover insightsthrough intelligent gap analysis and pattern recognition
This would transform Recall from a basic storage tool into an
intelligent knowledge exploration platform
- exactly what InfraNodus demonstrates is possible. The automatic relationship detection capability would be particularly transformative, eliminating Recall's current dependence on manual connections and creating a truly autonomous knowledge mapping system.Reference
: - Advanced AI-powered text analysis and knowledge graph platform for research and exploration.Sankari Nair
Merged in a post:
Please make the graph view more compact!
2
2voo1z4fn
The graph view shows the notes sometimes with such a distance to each other that it is way to complicated to look at them...
Sankari Nair
Merged in a post:
Graph view on the iPhone
L
Lana Droll
This feature would allow users to access the valuable graph view on their iPhones without needing to switch to a Mac.
Sankari Nair
Merged in a post:
Enhancing Exploration with Color-Coded and Labeled Graph Views
C
Carter Krech
As a Recall user,
I want an updated graph view feature that includes color coding and visible names for the nodes in both 2D and 3D views,
So that I can more easily distinguish between different types of knowledge cards, understand their relationships at a glance, and navigate my knowledge base more intuitively.
Color Coding: Users can toggle on color coding for nodes in the graph view, with colors representing different tags, link types, or other categorizations, making it easier to visually distinguish between them.
Floating Names: Option to toggle on visible names for each node without needing to hover, allowing users to quickly identify knowledge cards in the graph.
Customization Options: Users have control over the color schemes and which tags or types correspond to specific colors, allowing for a personalized visualization experience.
Toggle Features: Easy-to-use toggles are available for turning on or off the color coding and floating names features, enabling users to choose the view that best suits their current need.
Performance Optimization: The updated graph view is optimized for performance, ensuring smooth navigation and interaction, even with a large number of nodes.
User Preferences: Settings for these visualization features can be saved as part of the user's preferences, so they don't have to be reconfigured with each session.
Enhanced Navigation Tools: Alongside color coding and floating names, additional navigation tools (such as search within the graph) are enhanced for better usability.
Sankari Nair
Merged in a post:
Search and Highlights Function for Graph Section
j
joe faizano
User should able to key in Seach Function in Graph Section whereby it can search at the same time highlights or any area in the Graph that the key word is exist
Sankari Nair
Merged in a post:
Uncategorized cards in the Knowledge Graph
B
Benjamin Harrison (c0nsilience)
Being able to filter for 'uncategorized' cards in the knowledge graph would help with keeping it tidy.
Sankari Nair
Merged in a post:
Different colours for nodes in Graph view
M
Michael Killen
I love graph view, but sometimes when I'm using it I want to explore I can see the coloured lines. But it would be great to know if the "dots" are tags, actual content, or topics. I'd love to be able to colour code them so I can see or even filter what I'm looking at
Load More
→