jacktol commited on
Commit
84b30a1
1 Parent(s): 5e8b843

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +27 -0
README.md CHANGED
@@ -243,3 +243,30 @@ configs:
243
  - split: train
244
  path: data/train-*
245
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
243
  - split: train
244
  path: data/train-*
245
  ---
246
+
247
+ # USDA Branded Food Dataset
248
+
249
+ The USDA Branded Food Dataset is a cleaned and structured dataset created from USDA FoodData Central (FDC) data. This dataset consolidates extensive information on branded food items available in the U.S., providing a valuable resource for nutrition analysis, machine learning, and consumer research.
250
+
251
+ ## Overview
252
+
253
+ The USDA Branded Food Dataset compiles data on over 456,000 branded food items, including:
254
+ - **Nutrient Profiles**: Detailed nutrient content for each item.
255
+ - **Ingredient Lists**: Comprehensive ingredient information.
256
+ - **Serving Sizes**: Standardized serving size information for analysis and comparison.
257
+
258
+ By transforming and standardizing multiple data files from USDA FoodData Central, this dataset offers an accessible, machine-learning-ready resource for various food and nutrition applications.
259
+
260
+ For more information on how this data pipeline was developed, please refer to the detailed blog post [here](https://jacktol.net/posts/building_a_data_pipeline_for_usda_fooddata_central/).
261
+
262
+ ## Try the USDA Food Assistant
263
+
264
+ You can interact with the USDA Branded Food Dataset using the USDA Food Assistant, an application that allows for conversational exploration of food data. Try it on Hugging Face Spaces [here](https://huggingface.co/spaces/jacktol/usda-food-assistant).
265
+
266
+ ## See the Code
267
+
268
+ The code for the data pipeline that created this dataset, as well as the USDA Food Assistant, is available on GitHub [here](https://github.com/jack-tol/usda-food-assistant).
269
+
270
+ ## License
271
+
272
+ This dataset is shared under the MIT License.