Infoscape

The classification layer behind Skysquare — a domain and topic intelligence system for making public discourse easier to navigate.

Source Classification / IPTC Taxonomy / Skysquare Infrastructure
Active · Infrastructure Founder / engineer 2024–
Stack: Python · FastAPI · SQLAlchemy · PostgreSQL · pgqueuer · Docker Compose · curl_cffi → Camoufox → Wayback Machine (3-tier retrieval) · Fine-tuned classification model · IPTC-extended taxonomy

Infoscape is the classification layer behind Skysquare: a domain and topic intelligence system for making public discourse easier to navigate.

Its job is not to be a consumer product on its own. It exists to help Skysquare understand the web beneath the conversation — what kind of source a domain represents, what topics its posts belong to, and where authority or relevance may exist inside a large social graph.

The long-term goal is to map source domains and shared posts into an IPTC-aligned taxonomy, so readers can move through Skysquare by subject area, domain expertise, institutional type, and public relevance instead of being trapped in a flat stream of links.

Problem

A social graph is only useful if people can navigate it meaningfully.

Bluesky can show that people are sharing and discussing articles. But without source classification and topic mapping, those conversations remain hard to organize. A post about Taiwan, climate policy, semiconductor supply chains, Russian disinformation, local housing, or public health may all look like the same kind of link unless the system can understand what domain it came from and what subject area it belongs to.

That is the deeper problem Infoscape solves for Skysquare: not simply whether a source is “good” or “bad,” but what kind of source it is, what domain of knowledge it participates in, and how it should help a reader discover relevance and authority.

Concept

Infoscape turns the open web into a structured source map for Skysquare.

At the domain level, it verifies and classifies information sources: newspapers, universities, government agencies, NGOs, think tanks, advocacy groups, platforms, blogs, commercial sites, spam networks, propaganda outlets, and other recurring categories of web authority.

At the topic level, it is designed to map posts and sources into an IPTC-extended taxonomy, giving Skysquare a navigational layer beyond raw social activity. The goal is to let readers move through discourse by subject matter and authority signals: who is discussing a topic, which sources are being cited, what institutions are involved, and which conversations are clustered around a domain area.

The admin interface exists to support that work: review classifications, inspect confidence, validate edge cases, and improve the source map that powers the reader-facing Skysquare experience.

Why it matters

Public discourse is not just a stream of opinions. It is a map of sources, subjects, institutions, claims, expertise, and contested authority.

Skysquare shows the conversation around the page. Infoscape helps determine what kind of page it is, what topic space it belongs to, and how it should connect to adjacent conversations. That makes the social graph more useful to readers, researchers, journalists, analysts, educators, and anyone trying to follow a complex issue across the open web.

This is what makes Infoscape epistemic infrastructure: it gives Skysquare a way to organize discourse around source type, subject area, and authority rather than mere engagement.

Status

Infoscape is active infrastructure for Skysquare. It currently supports large-scale domain verification, scraping, classification, confidence scoring, and human review workflows.

The system is being extended toward IPTC-aligned topic mapping for posts and source domains, so Skysquare can support more useful discovery and filtering across public discourse.