← Project home · Live application · Windows EXE · Download PDF · DOI

Official visual companion

Inside GeoBusiness Intelligence Studio

A Visual Guide to Open Data, Geospatial Discovery, and Local Business Intelligence

GeoBusiness Intelligence Studio overview

Faramarz Kowsari · Version 1.2.0 · DOI 10.5281/zenodo.21539094

01

Purpose

Why build a local-business intelligence laboratory?

Business-location data often arrives fragmented: names vary, addresses are incomplete, contact fields disappear, and the same place may appear more than once. GeoBusiness Intelligence Studio converts that ambiguity into an inspectable pipeline.

02

Providers

One interface, three data paths

Offline fictional samples support predictable demonstrations. OpenStreetMap uses Nominatim and Overpass with safeguards. The optional Google Places adapter uses the official API and the user's own restricted key.

03

Workflow

From a query to an exportable dataset

Choose provider, business type, city, radius, and limit. Resolve geography, retrieve records, normalize fields, score completeness, remove probable duplicates, map results, summarize coverage, and export CSV or JSON.

04

Interface

A dashboard that keeps evidence visible

Search controls expose provider and geography. Metric cards reveal coverage and average quality. The map preserves spatial context, while tables keep source, contacts, ratings, and quality visible.

05

Architecture

FastAPI at the center of replaceable components

Provider adapters feed validated FastAPI routes. Search orchestration performs normalization, quality scoring, sorting, and conservative deduplication. Static HTML and Leaflet provide the browser interface.

06

Data quality

Completeness is measured; identity is inferred cautiously

The quality score measures the presence of useful fields, not factual correctness. Duplicate detection combines stronger evidence such as matching phone numbers with softer evidence such as similar names and addresses.

07

Privacy, AI, and deployment

Useful without AI; extensible when AI is justified

The deterministic workflow is complete without AI. Ollama and OpenAI-compatible endpoints are optional. The project supports Render, Docker, source execution, and a verified one-file Windows edition.

08

Use cases and citation

Research value with permanent distribution

Use the project for market orientation, data-engineering education, responsible-AI demonstrations, and portfolio review. Version 1.2.0 is preserved on Zenodo and identified by a permanent DOI.

09

Author

Faramarz Kowsari

Faramarz Kowsari is an author, Software Engineer and AI researcher based in Istanbul. Official profiles include ORCID, Google Scholar, GitHub, LinkedIn, and Google Books.

The map becomes intelligence only after the uncertainty remains visible.

10