HoodEx is a decision-support tool, not a prediction of how you will experience a place. It organises available evidence into comparable signals so you can build a shortlist, inspect trade-offs, and decide what to verify yourself.
A public area score runs from 0 to 10 and summarises measured neighbourhood conditions across HoodEx categories. It is attached to an area, not to a person.
A personal fit score is a weighted result from 0 to 100 based on the priorities you select. It is not a percentile and does not mean that an area is better than a stated percentage of other neighbourhoods. Two people can receive different fit scores for the same area.
HoodEx combines signals derived from satellite imagery, mapped features and amenities, and public research or datasets. Depending on the measure, a result may describe a count, density, distance, proximity, coverage, or modelled condition.
A nearby or numerous feature is not automatically a high-quality feature. Counts and proximity should not be read as ratings of service quality, accessibility, opening hours, capacity, or personal suitability.
Housing-cost indicators describe available area-level costs per square metre. They are not live property listings, a valuation of a particular home, or a guarantee that housing at that price is available.
Sources update on different schedules and geographic coverage is not uniform. A missing feature may mean that it is absent, not mapped, unavailable from a source, or outside the period represented by the data.
Some values are modelled or estimated when direct observations are limited. Low-confidence estimates and incomplete coverage deserve extra caution; comparisons are strongest when the underlying coverage is similar.
Crime and safety data can differ in reporting practice, geography, time period, and completeness. Recorded incidents do not capture every event, and an area-level signal cannot predict personal safety at a specific street or time.
Never treat a score as a safety guarantee. Consult current official local guidance, consider time-of-day and route conditions, visit where practical, and use your own judgment.
Selected priorities contribute to a weighted personal fit score. A commute requirement can also act as a hard constraint: an otherwise appealing area may be excluded when it does not meet that requirement.
Property preferences that are not part of the neighbourhood scoring engine are shown as context rather than used to rank areas. The match is therefore a neighbourhood shortlist, not a claim that a suitable property is available.
Local scenario travel values are approximations for comparing areas. They are not real routed trips and may not reflect the exact street network, schedules, traffic, transfers, gradients, disruptions, or accessibility needs.
Verify important journeys with a current routing or transport service before making a decision.
AI-generated interpretation can help summarise and explore the underlying results, but it does not create the source observations. It can omit context or be wrong.
Use cited or displayed evidence for factual claims, and verify consequential details with authoritative sources. An AI explanation does not increase the confidence or freshness of the underlying data.
Every area is scored from 0 to 10 in five categories. They come from Ekistics, a framework for studying human settlements, and the original names are listed for anyone reading the research.