README.md

metadata

language:
  - en
license: other
pretty_name: 'Personal Query: Pet Supplies'
task_categories:
  - text-generation
  - text-retrieval
tags:
  - personalized-search
  - query-generation
  - error-query
  - synthetic-data
  - amazon-reviews
configs:
  - config_name: full
    data_files:
      - split: train
        path: data.jsonl
  - config_name: paired
    data_files:
      - split: train
        path: paired_data.jsonl

Personal Query: Pet Supplies

This dataset contains personalized product search queries for the Pet_Supplies category.

Each record is built from the Personal Query pipeline:

• Stage 6 generated correct personalized queries.
• Stage 7 injected user-specific error query variants when a matching error pattern was available.
• Stage 5 provided the user profile complexity level.

Files

• data.jsonl: all correct Stage 6 queries. Rows without Stage 7 error query keep error_query as null.
• paired_data.jsonl: only rows where a correct query has a paired error query.
• summary.json: generation statistics for this category.

Dataset Size

Metric	Count
Total rows in data.jsonl	40814
Rows in paired_data.jsonl	2589
Rows without noisy query	38225

Rows By Query Category

query_category	rows
deep	20407
wide	20407

Paired Rows By Query Category

query_category	rows
deep	1954
wide	635

Rows By Complexity

query_category:level	rows
deep:0	4293
deep:1	9288
deep:2	4761
deep:3	2065
wide:0	9225
wide:1	8423
wide:2	2276
wide:3	483

Schema

Field	Type	Description
category	string	Product category.
uuid	string	User identifier.
asin	string	Amazon product identifier.
query_category	string	Query type: wide or deep.
complexity_level	integer	Complexity level of the generated Stage 6 query.
profile_complexity_level	integer	User profile complexity level from Stage 5.
correct_query	string	Correct query generated in Stage 6.
correct_word_count	integer	Word count of correct_query.
attrs_used	object	Product attributes used to generate the query.
has_error_query	boolean	Whether a Stage 7 error query is available.
error_query	string or null	Error query generated in Stage 7.
injected_errors	list	Injected error metadata for error query generation.

Loading

from datasets import load_dataset

full = load_dataset("xxxxdszz/personal-query-pet-supplies", name="full", split="train")
paired = load_dataset("xxxxdszz/personal-query-pet-supplies", name="paired", split="train")

Intended Use

This dataset is intended for research on personalized product search, query generation, error query robustness, and retrieval evaluation.

Data Notes

The queries are synthetic outputs generated from user/product signals in the local Personal Query pipeline. The error queries are generated by injecting user-specific writing error patterns. Review license and redistribution requirements for any upstream source data before external reuse.