The publicly available open GreenDB is a product-by-product sustainability database. It contains product attributes, e.g., name, description, color, etc. and, on the other hand, information about the products' sustainability. Moreover, the sustainability information is transparently evaluated so that it is possible to rank products depending on their sustainability. Landing Page: https://green-db.calgo-lab.de Source Code: https://github.com/calgo-lab/green-db Further Notes For usability, we export two datatypes: CSV Parquet (using pyarrow python package) Be careful with the columns' data types when using the CSV files! Use for example: from ast import literal_eval from datetime import datetime import pandas as pd products = pd.read_csv("products.csv", converters={"timestamp": datetime.fromisoformat, "image_urls": literal_eval, "sustainability_labels": literal_eval}).convert_dtypes() sustainability_labels = pd.read_csv("sustainability_labels.csv", converters={"timestamp": datetime.fromisoformat}).convert_dtypes()

{"references": ["J\u00e4ger, S., Greene, J., Jakob, M., Korenke, R., Santarius, T., & Biessmann, F. (2022). GreenDB: Toward a Product-by-Product Sustainability Database. ArXiv, abs/2205.02908.", "Flick, A., J\u00e4ger, S., Garcia, J.A., Driesch, K.V., Brendel, K., & Biessmann, F. (2022). GreenDB - A Dataset and Benchmark for Extraction of Sustainability Information of Consumer Goods. ArXiv, abs/2207.10733."]}

Supported by the Federal Ministry for the Environment, Nature Conservation and Nuclear Safety based on a decision of the German Bundestag. Förderkennzeichen: 67KI2022B