Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.snorbe.deskrex.ai/llms.txt

Use this file to discover all available pages before exploring further.

Knowledge Graph

The Knowledge Graph is a visual “map” of research results that Snorbe has collected and analyzed. Circular dots (nodes) represent units of information, while lines represent connections between them. In short, it’s like a bird’s-eye view diagram of all your research.

Understanding the Graph View

When you open the Knowledge Graph, you’ll see numerous circular dots connected by lines. It might look complex at first glance, but the concept is simple.

Circular Dots (Nodes)

Units of information handled in your research, such as companies, technologies, papers, patents, and people. Colors represent the type of information.

Lines (Edges)

Relationships between pieces of information. The type of connection—“related to,” “cites,” “competes with”—depends on the context.

Information Types (Node Colors)

Node colors indicate what type of information they represent.
ColorInformation Type
Light BlueAgent Run (research unit)
GreenPublic Source (web information)
OrangePrivate Source (uploaded files)
GrayEntity (information unit)
The small dot in the center of each node represents the community (group of related information) that the node belongs to. Information that is strongly related to each other is grouped together as a community. For example, a community about “semiconductor materials” might include material manufacturers, related patents, and application technologies. Each group is color-coded, so you can see at a glance “how much information exists in each field.” You can also use the community filter to display only specific groups. Community boundaries are displayed as ellipses. As information grows, the boundary expands, making it easier to visually grasp the outline of each group.

Search Functionality

When searching for information in the Knowledge Graph, two methods are combined.

Keyword Search

Infers related terms from your input keywords and searches for entities (names, companies, technical terms, etc.) on the graph.

Vector Search

Searches by semantic similarity. Finds information that is close in meaning even if the words are different, like “apple” and “fruit.”
By running both simultaneously, you can perform broad searches that include spelling variations and related terms.

Basic Operations

You can perform the following operations in the graph view.

Zoom

Use the mouse wheel or pinch gestures to zoom in and out. Zoom out to see the big picture, zoom in for details.

Pan

Drag the graph to move the area you want to see to the center of the screen.

Select Node

Click on a circular dot to display detailed information about it in the side panel.

View Relationships

When you select a node, other information connected to it is highlighted.

Click a Node to See Details

Click on a circular dot that interests you, and a detail panel will open on the right. Here you can see:
  • Name and type of information
  • Summary and description
  • Related information (list of directly connected entities)
  • Research history (which agent run collected this information)
  • References to source data (DOI for papers, URLs for companies, etc.)
Press the zoom button in the top-right corner of the detail panel to move the camera to that node’s location on the graph. This is useful when you’re viewing the big picture and want to find where a specific entity is located.

Switching Between 2D and 3D

The graph can be switched between 2D and 3D display. In 3D mode, clusters of information appear in three dimensions, making it easier to grasp the overall structure of large graphs.

Auto-Arranging Nodes

When the graph becomes complex, use the alignment feature to automatically rearrange nodes. Grouping them by entity type (people, organizations, locations, etc.) makes the overall layout much easier to read.

Layout is Saved

After you rearrange nodes on the graph, the same layout is preserved when you open it next time. Once organized, you don’t need to redo it every session. The “find similar entities” feature automatically discovers other entities with characteristics similar to the one you’re currently focused on. Example uses:
  • In sales planning, find “Municipality B with similar challenges to Municipality A”
  • In IP strategy, identify “companies with technology areas similar to a competitor”
  • In R&D, discover “related technology trends”
The agent automatically combines graph database queries and vector search to extract highly relevant information.

Easy Fact-Checking

Snorbe saves the original data from web pages and documents referenced during research. This makes it easy to trace back which source any statement in a report is based on.

Source Transparency

Access the original web page or document directly from each node or report entry.

Reuse of Past Data

Information you’ve already researched is stored as embeddings, so you don’t need to repeat the same queries.

White Space Detection

The most powerful feature of the Knowledge Graph is its ability to discover unexplored areas (white spaces).

How Unexplored Areas Appear

White spaces appear on the graph as “sparse areas with little information” or “completely blank areas.”

Sparse Areas

Areas where dots are widely spaced and lines are few. Related information is fragmented, requiring deeper investigation.

Blank Areas

Areas with no information at all. Areas expected to be related to your research topic but where no data exists.
Snorbe automatically detects these areas and suggests them as “candidates for your next research.”

Why White Spaces Become Visible

With traditional research methods, you review information in list format, making it easy to miss “what’s missing” from the whole picture. The Knowledge Graph visualizes information density, making it easier to notice gaps.

Benefits

Using the Knowledge Graph provides the following advantages:

Grasp the Big Picture

See at a glance how comprehensive your collected information is. Research progress is also visualized.

Prevent Oversights

Discover “information gaps” that are hard to notice in list format, preventing important oversights.

Discover New Connections

Find unexpected relationships between pieces of information, leading to new hypotheses and ideas.

Common Use Cases

Prevent Research Gaps in Competitive Analysis

When researching competitor technology trends, use the Knowledge Graph to get an overview and discover companies or technology areas you might have missed. Example: While researching battery materials, you had identified major manufacturers, but the Knowledge Graph revealed the presence of university spin-offs.

Determine Research Direction with Technology Mapping

Get an overview of a technology field and identify which areas are unexplored to determine your R&D direction. Example: In next-generation semiconductor research, you noticed that “materials development” had plenty of information but “manufacturing processes” was thin, so you decided to dig deeper there.

Discover New Business Opportunities

Visualize white spaces in the market to discover areas where competitors haven’t ventured. Example: In healthcare research, you discovered little information at the intersection of “for seniors” and “wearable,” leading to new product ideas.

Back to Features

Explore other Snorbe features