A research program needs a venue map, not one favorite feed

A research program becomes taste-driven when the team keeps returning to the same familiar feed and starts calling that habit coverage. The lesson is simple: ontology work does not live in one publication lane. Useful advances show up across ontology and Semantic Web venues, logic and reasoning communities, application papers, and domain-specific outlets where the modeling problem matters more than the ontology label.

Hadto is trying to turn employees into business owners, so this problem reaches beyond academia. A venture-building research loop has to become something a later operator can run, challenge, and continue. The learning system is still personal judgment disguised as process when staying current depends on one insider knowing which journal, workshop, or mailing-list thread “really matters.”

A narrow feed becomes private judgment

A founder can compensate for a narrow intake loop for quite a while. They already know which sources are missing, so they mentally patch the gaps and quietly keep one eye on the unofficial extras. The successor does not inherit that instinct. They inherit the visible system, and a visible system that points to one preferred venue family quietly turns ownership back into private memory.

The real risk is operational, not academic. A narrow research feed makes the business dependent on whoever knows the missing map.

What a venue map should actually contain

Ontology engineering sits in an awkward but useful position. The work appears in formal ontology venues, Semantic Web conferences, logic communities, tooling discussions, and application-heavy domains where the paper may never use the word “ontology” in the title. No single slice captures the whole picture.

A usable venue map for Hadto might separate formal ontology and Semantic Web venues for modeling discipline, reasoning and logic venues for constraint handling, tooling venues for implementation patterns, and application-heavy outlets where ontology choices are tested inside real workflows. The same map should tag maturity signals: workshops and posters as early indicators worth watching, conference papers as patterns worth testing, and journals or survey articles as candidates for operating defaults. It should also record deliberate coverage gaps. A cycle focused on adoption mechanics may skip highly theoretical description-logic work for that pass. Vendor blog posts without implementation detail can be logged but kept out of decision-making. Those omissions belong on the map too, because the next operator needs to know whether silence means “not relevant this round” or simply “not checked.”

Publication type should change the action

Venue families are not the only distinction. Publication types differ too. An early workshop signal may justify adding a question to the watchlist or running a small internal test. A repeated conference pattern may justify prototyping a new review step. A journal synthesis may be strong enough to change a default method or update shared venture infrastructure. Early signals and mature results are both useful. The mistake is pretending they should carry the same decision weight.

What a successor should be able to inspect

An owner-operator does not need a research program that merely sounds smart. They need one that shows where the team looks, why those sources belong on the watchlist, what kind of maturity each source tends to represent, which gaps were chosen on purpose, and how that evidence should influence action. The same principle shows up outside ontology work. A company should not rely on one salesperson’s memory of where market signals surface, one operator’s private sense of which customer complaints matter, or one technical founder’s personal map of the field.

When the watchlist becomes part of the company’s standing research routine, each reading cycle becomes an auditable asset instead of a taste profile.

Operating standard

The standard is straightforward: if a successor cannot inspect the venue map and tell what is covered, what is intentionally out of scope, and how different signals change decisions, the company does not yet have a research process another operator can inherit. It has a habit.

Hadto should treat the venue map as required operating infrastructure and keep it reviewable alongside the findings it produces. When a company says it is learning from the field, that claim should survive handoff without requiring the next operator to reconstruct the missing map from someone else’s memory.

---

Source evidence used in this note: reviewed internal Keet study materials and current ontology research-program operations guidance (reviewed 2026-04-20), along with existing Hadto blog posts checked to avoid duplicating earlier notes on evidence attachment, publication readiness, and discovery coverage.