This dataset contains property sales data, including information such as PropertyID, property type (e.g., Commercial or Residential), tax keys, property addresses, architectural styles, exterior wall materials, number of stories, year built, room counts, finished square footage, units (e.g., apartments), bedroom and bathroom counts, lot sizes, sale dates, and sale prices. Explore this dataset to gain insights into real estate trends and property characteristics.

Data.milwaukee.gov, (2023). Property Sales Data. [online] Available at: https://data.milwaukee.gov [Accessed 9th October 2023].

Open Definition. (n.d.). Creative Commons Attribution 4.0 International Public License (CC BY 4.0). [online] Available at: http://www.opendefinition.org/licenses/cc-by [Accessed 9th October 2023].

155M U.S. property records with ownership, valuation, assessment, and characteristic data. Residential + commercial across all 50 states, including HI, MT, WI.

Sold transaction records for 8,523 Ohio residential properties across 5 property types and 814 ZIP codes. Includes list vs sold price, property specs, negotiation outcome (Above / At / Below Asking), sold-to-list ratio, price premium, year built, and PII-redacted listing descriptions for 8,402 records.

APPLICATIONS - Sale price prediction: model lastSoldPrice from sqft, beds, baths, type, garage, year_built, and postal_code using regression. - Negotiation outcome classification: predict Above / At / Below Asking result using property features, type, and price bracket as inputs. - Market zone analysis: compare sold-to-list ratios and negotiation outcomes by ZIP code to identify buyer's vs seller's market pockets across Ohio. - NLP / text mining: extract amenity keywords from 8,402 PII-redacted listing descriptions and correlate with price premium to score listing language value. - Historical stock analysis: use year_built (1800-2026) to study how property age affects pricing, negotiation outcomes, and market demand across Ohio regions.

COLUMNS (VARIABLES) - type string Property type: single_family, condos, townhomes, multi_family, other - sub_type string Property sub-type (100% null - structural, no sub-type data available in source) - listPrice float Original asking price, USD - sqft float Interior square footage - stories float Number of storeys (2.0% null) - beds float Number of bedrooms (note: multi_family records may reflect total unit count, not individual beds) - baths float Total bath count - baths_full float Full bathrooms only - baths_full_calc float Calculated full bath count from source data - garage float Garage capacity in car spaces (15.8% null = no garage data recorded) - lastSoldPrice int Final recorded sale price, USD (primary target variable; fully populated, 0 nulls) - postal_code float 5-digit Ohio ZIP code of the property - is_valid_oh_zip bool True if postal_code is a confirmed Ohio ZIP (all 8,523 rows = True) - zip_candidate_count int Number of Ohio ZIP patterns matched in source URL - sold_to_list_ratio float lastSoldPrice / listPrice - price_premium_pct float Percentage premium or discount vs list price (positive = sold above asking, negative = below) - negotiation_outcome string Above Asking / At Asking / Below Asking (derived from sold_to_list_ratio; 0 Unknown rows) - ratio_outlier_flag bool True if sold_to_list_ratio is an IQR outlier (814 rows flagged; not removed) - text_clean string Property listing description with PII redacted (available for 8,402 of 8,523 records) - pii_redacted_flag bool True if PII was found and redacted within text_clean (1,383 records) - year_built float Year the property was constructed (1800-2026; 69 nulls; Ohio-specific bonus column not present in companion SC dataset)

IMAGE CREDITS Image generated using ChatGPT (OpenAI).

LICENSE CC BY-NC-SA 4.0 (Attribution-NonCommercial-ShareAlike)

Records are derived from publicly listed property data with all PII removed. Published for research and educational use. Users should independently verify compliance with applicable MLS data terms for any commercial application.

Graph and download economic data for Real Residential Property Prices for United States (QUSR628BIS) from Q1 1970 to Q1 2026 about residential, HPI, housing, real, price index, indexes, price, and USA.

The ProspectNow Data API delivers all the data and metadata you need for residential and commercial properties across the U.S.

It is designed to provide flexibility, as well as qualified, up-to-date data from a dependable source, so you can focus on providing great customer experiences.

Whether you want to enrich existing datasets, improve your own customer-facing application, or integrate our data into your tech stack, we have everything you need in our REST API, including:

Property Ownership Building Characteristics Valuation Mortgage Information Foreclosure/Preforeclosures Property Tax Info Market Data Properties Predicted to Sell Properties Predicted To Refinance +more

Discover the booming real estate & property management services market! This in-depth analysis reveals key trends, growth drivers, and challenges impacting the industry from 2019-2033, including insights on leading companies and regional performance. Learn about the lucrative opportunities and potential risks in this dynamic sector.

This dataset represents real estate assessment and sales data made available by the Office of the Real Estate Assessor. This dataset contains information for properties in the city, including acreage, square footage, GPIN, street address, year built, current land value, current improvement value, and current total value. The information is obtained from the Office of the Real Estate Assessor ProVal records database. This dataset is updated daily.

Comprehensive property data, parcel information, and ownership records for Sacramento County, CA. Access real estate market insights and property details.

Title Arizona Real Estate: Sold Properties Dataset 2026

Subtitle 9,954 verified closed deals with price, sqft, ZIP & summarized text

Dataset Overview

This dataset captures 9,954 verified sold residential properties across the state of Arizona, covering 292 unique ZIP codes spanning the full 85xxx and 86xxx range — from Phoenix and Scottsdale to Sedona, Flagstaff, and rural Maricopa County.

Every record is a confirmed closed transaction, not an active listing or automated estimate. Combined closed sales volume stands at $6.52 billion, with a median sold price of $459,928 and a range from entry-level homes under $50K to luxury estates exceeding $21 million.

All listing descriptions have been processed through a two-stage anonymization pipeline — a PII redaction engine strips emails, phone numbers, street addresses, agent names, MLS IDs, and location identifiers, followed by GPU-accelerated abstractive summarization to produce clean, paraphrased descriptions capped at 50 words with meaningful sentence endings. All numerical fields have been normalised and listing descriptions AI-paraphrased to ensure no individual property can be identified from this dataset alone.

Two analyst-ready derived features are included out of the box: price_per_sqft and list_to_sold_ratio. The latter reveals that 30.3% of Arizona properties sold at or above asking price — a direct market heat signal embedded in every row.

Data Science Applications

• Automated Valuation Models (AVM) — train regression or gradient-boosted models on closed price using ZIP, sqft, beds, baths, year built, and garage
• Market Heat Analysis — list_to_sold_ratio maps over- and under-heated ZIP codes across the full Arizona market
• NLP & LLM Fine-tuning — sanitized_text delivers abstractive, domain-specific real estate prose ready for embedding or generative fine-tuning
• Geospatial Intelligence — 292 ZIP codes pair directly with US Census TIGER shapefiles for choropleth market mapping
• Price Premium Modelling — luxury keyword signals in sanitized text correlate with sold price premium
• Privacy-Preserving ML Research — normalised values and paraphrased text make this a clean benchmark for privacy-aware modelling

Column Descriptors

Provenance

Source: Publicly available residential real estate listing data. All records correspond to properties with a confirmed sold status at time of collection.

Methodology: Listings were collected via structured web extraction targeting closed-sale records only. ZIP codes were parsed from standardised listing URL slugs. A multi-rule PII redaction pipeline was applied covering emails, phone numbers, street addresses, agent names, MLS and parcel IDs, HOA fees, and Arizona-specific location identifiers. Redacted text was then passed through GPU-accelerated abstractive summarization to produce paraphrased 50-word property descriptions. Structural fields were validated against physical bounds and outliers nulled rather than imputed to preserve analytical integrity.

Acknowledgements

Dataset thumbnail generated with Google ImageFX — more state-level closed sales datasets are in development.

Tags

real estate · arizona · housing prices · nlp · regression

Graph and download economic data for Commercial Real Estate Prices for United States (COMREPUSQ159N) from Q1 2005 to Q2 2025 about real estate, commercial, rate, and USA.

This dataset provides detailed information about property sales, including various property features and sale statistics. The data spans multiple years and includes information about towns, property types, sale amounts, assessed values, and additional remarks from assessors. Below is an overview of the key columns:

1-Serial Number: A unique identifier for each property record.

2-List Year: The year when the property was listed for sale.

3-Date Recorded: The date the property sale was recorded in the dataset.

4-Town: The town or city where the property is located.

5-Address: The street address of the property.

6-Assessed Value: The value assigned to the property for tax purposes.

7-Sale Amount: The final sale price of the property.

8-Sales Ratio: A ratio of the sale amount to the assessed value, potentially indicating how close the sale price is to the assessed value.

9-Property Type: The type of property (e.g., Residential, Commercial).

10-Residential Type: The specific type of residential property, such as Single Family.

11-Non Use Code: Code indicating properties that may not be used for typical purposes (e.g., vacant land).

12-Assessor Remarks: Additional remarks from the assessor about the property.

13-OPM Remarks: Remarks from the Office of Property Management.

14-Location: Geographical coordinates of the property (latitude and longitude).

Labels and Categories: Price Ranges: Several columns categorize properties into different price ranges based on their sale amount, assessed value, etc.

Time Periods: Data is also grouped by time periods (e.g., different months or years), with counts of properties that fall within each time range.

Location: Some rows include coordinates indicating the exact location of the property (longitude and latitude).

Use Cases: Regression:

You can predict Sale Amount based on other features like Assessed Value, Sales Ratio, Property Type, and Location.

Classification:

Classify properties based on Property Type or Residential Type.

You can also create a classification model to predict whether the Sales Ratio falls within a specific range.

This dataset could be particularly useful for real estate analysis, pricing models, or exploring patterns in property sales over time.

Comprehensive real estate market analysis for Northfield, Illinois, covering sales trends, pricing, and market conditions.

ATTOM’s Vacancy Data delivers a nationwide view of long-term residential vacancy, identifying properties that have been vacant for at least 90 days across the United States. This dataset combines Property Owner Data, Rental Data, Real Estate Market Data, and Residential Real Estate Data to support ownership analysis, market insight, and opportunity identification.

Each record includes verified, standardized address-level information and detailed ownership data, allowing organizations to accurately locate vacant residential properties and understand ownership patterns at scale. Updated monthly, the dataset reflects current market conditions and captures vacancy trends across counties and ZIP codes nationwide.

By highlighting underutilized residential assets, Vacancy Data enables users to analyze vacancy dynamics, remove inactive addresses from outreach efforts, and identify properties that may represent strategic opportunities for investment, redevelopment, or targeted services.

Comprehensive property data, parcel information, and ownership records for Palm Beach County, FL. Access real estate market insights and property details.

This dataset contains over 16K+ property listings from zameen.com, a prominent online property portal in Pakistan. It includes detailed information on each property, such as city, location, price in PKR, number of bedrooms and bathrooms, and property size in square feet. This comprehensive dataset is a valuable resource for real estate analysts and professionals seeking to explore the Pakistani housing market. The data can be utilized for market and trend analysis, investment research, and other related purposes.

This data is scrapped using the zameen-com-scrapper.

DATASET IN BRIEF Sold transaction records for 7,805 Alabama residential properties across 6 property types and 421 ZIP codes. Includes list vs sold price, property specs, negotiation outcome (Above / At / Below Asking), sold-to-list ratio, price per square foot, property age, sale date features, and PII-redacted listing descriptions for 6,977 records.

APPLICATIONS - Sale price prediction: model lastSoldPrice from sqft, beds, baths, property type, garage count, property_age, and postal_code using regression. - Negotiation outcome classification: predict Above / At / Below Asking result using property features, type, and price bracket as inputs. - Market zone analysis: compare sold-to-list ratios and price_per_sqft by ZIP code to identify buyer's vs seller's market pockets across Alabama. - Seasonality research: use sold_month and sold_quarter to identify temporal patterns in sale prices and negotiation outcomes. - NLP / text mining: extract amenity keywords from 6,977 PII-redacted listing descriptions and correlate with price premium to score listing language value.

COLUMNS (VARIABLES) - type string Property type: single_family, condos, townhomes, multi_family, other, duplex_triplex - sub_type string Property sub-type where available (93% null for single-family records which carry no sub-type) - listPrice float Original asking price, USD (null where not publicly disclosed; ~10% of records) - lastSoldPrice float Final recorded sale price, USD (primary target variable; 100% populated) - soldOn string Date of sale, YYYY-MM-DD format - sqft float Interior square footage (100% populated) - stories float Number of storeys - beds float Number of bedrooms (100% populated) - baths float Total bath count - baths_full float Full bathrooms only - baths_full_calc float Calculated full bath count (from source data) - garage float Garage capacity in car spaces (null = no garage data; ~46% of records) - year_built float Year of original construction - postal_code string 5-digit ZIP code of the property - is_valid_al_zip bool True if postal_code is a confirmed Alabama ZIP (range 35004-36925) - sold_year float Year component extracted from soldOn - sold_month float Month component extracted from soldOn (1-12) - sold_quarter float Quarter component extracted from soldOn (1-4) - property_age float Years since construction (2026 minus year_built) - sold_to_list_ratio float lastSoldPrice / listPrice (null where listPrice not disclosed; 806 records) - price_premium_pct float Percentage premium or discount vs list price (positive = sold above asking, negative = below) - negotiation_outcome string Above Asking / At Asking / Below Asking / Unknown (derived from sold_to_list_ratio thresholds) - ratio_outlier_flag bool True if sold_to_list_ratio < 0.5 or > 1.5 (20 rows flagged; not removed) - price_per_sqft float lastSoldPrice / sqft, rounded to 2 decimal places (key comparables and valuation metric) - text_clean string Property listing description with PII redacted (available for 6,977 of 7,805 records) - pii_redacted_flag bool True if PII pattern-matching found and redacted content within text_clean (657 records)

IMAGE CREDITS Image generated using ChatGPT (OpenAI).

LICENSE CC BY-NC-SA 4.0 (Attribution-NonCommercial-ShareAlike)

Records are derived from publicly listed property data with all PII removed. Published for research and educational use. Users should independently verify compliance with applicable listing platform data terms for any commercial application.

Property addresses appearing in the url field reflect publicly recorded sold transaction data available through US county assessor and MLS public records. pii_redacted_flag applies to the text_clean field, where inline addresses and contact details were removed. Public-record addresses in url are retained by design

Property addresses appearing in the url field reflect publicly recorded sold transaction data available through US county assessor and MLS public records. pii_redacted_flag applies to the text_clean field, where inline addresses and contact details were removed. Public-record addresses in url are retained by design

Overview

This dataset provides a comprehensive snapshot of the Texas real estate market as of 2024, featuring a curated selection of 500 property listings. It encompasses a wide array of properties, reflecting the diverse real estate landscape across Texas. This dataset serves as a foundational tool for understanding market dynamics, property valuations, and regional housing trends within the state.

Data Science Application of Dataset

Given its breadth and depth, this dataset is poised to facilitate a multitude of data science applications. Researchers and analysts can leverage this dataset for exploratory data analysis (EDA) to identify patterns, trends, and anomalies within the Texas real estate market. It is particularly suited for regression analyses to predict property prices based on various features, classification tasks to categorize properties into different market segments, and geographical data analysis to understand regional market dynamics. Despite the dataset's modest size, it offers a rich source for machine learning models aimed at providing insights into price determinants and market trends, ensuring practical applications remain within realistic and achievable bounds.

Full Column Descriptors

• url: Web address for the property listing on Realtor.com.
• status: Current status of the listing, indicating availability.
• id: Unique identifier for each property listing.
• listPrice: The asking price for the property.
• baths: Total number of bathrooms, including partials.
• baths_full: Number of full bathrooms.
• baths_full_calc: Calculated number of full bathrooms, for consistency.
• beds: Number of bedrooms in the property.
• sqft: Total square footage of the property.
• stories: Number of levels or floors in the property.
• sub_type: Specific sub-category of the property, if applicable.
• text: Descriptive narrative provided for the property listing.
• type: General category of the property (e.g., single-family, condo).
• year_built: Year the property was constructed.

Ethically Mined Publicly Available Data Only

This dataset has been meticulously compiled, adhering to ethical standards and ensuring all data is sourced from publicly available information. It respects privacy and copyright considerations, utilizing data that is openly accessible and intended for public consumption.

Acknowledgments

Gratitude is extended to Realtor.com for serving as an invaluable resource in the compilation of this dataset. The platform's commitment to providing comprehensive and accessible real estate data has significantly contributed to the depth and quality of this dataset.

Image Acknowledgment

The dataset thumbnail image is credited to Realtor.com, as featured on their official Facebook page. The image serves as a visual representation of the diverse and dynamic nature of the Texas real estate market, captured in this comprehensive dataset. View Image

Verified closed sales data for Concord NC from Canopy MLS. Median sale price, average sale price, days on market, sale-to-list ratio, and inventory depth for all property types in Concord NC Cabarrus County.

The booming Real Estate Asset Management Software market is projected to surpass $10 billion by 2033, driven by cloud adoption and increased demand for efficient property management. Explore market trends, key players, and regional insights in our comprehensive analysis.

Wholesale Real Estate property and owner data for Tuscaloosa County, Alabama (FIPS 01125), sourced from Tuscaloosa County parcel layer with full data lineage.