bhmarketer/techbullion-faq-dataset

TechBullion Press Release FAQ Dataset

Data behind the BHMarketer TechBullion FAQ Generator, a question search tool that takes a question in plain English and returns the matching official answer from 40 TechBullion press release FAQs with a 0 to 100 confidence score. Built by BHMarketer.ai.

Prices and package details can change. Check the order page for the current offer. Not affiliated with TechBullion.

Files

File	Rows	Contents
techbullion_faqs.csv	40	FAQ knowledge base: id, category, question, answer, keywords
techbullion_search_benchmarks.csv	40	Real-world questions with the expected FAQ, confidence and band
techbullion_faq.py		The dependency-free Python search tool, so results can be reproduced

FAQ Columns

Column	Description
id	FAQ number, 1 to 40
category	One of service, buying, crypto, presale
question	The official question
answer	The official answer
keywords	Space-separated search keywords

Benchmark Columns

Column	Description
case_id	Row number
query	A question as a real user might type it
expected_id	The FAQ that should be returned
expected_question	That FAQ's question
confidence	Match score from 0 to 100
band	No match, Possible match, Good match or Strong match

Usage

from datasets import load_dataset

faqs = load_dataset("bhmarketer/techbullion-faq-dataset", "faqs")["train"]
bench = load_dataset("bhmarketer/techbullion-faq-dataset", "benchmarks")["train"]
print(faqs[11]["question"], faqs[11]["answer"])

import techbullion_faq as tb
print(tb.ask("can i pay with paypal"))

Links

Platform	URL
Service	https://bhmarketer.ai/products/techbullion/
GitHub	https://github.com/bhmarketer-ai/BHMarketer-buy-techbullion-press-release-faq-generator
NPM	https://www.npmjs.com/package/@bhmarketer-ai/techbullion-faq-generator
Kaggle	https://www.kaggle.com/datasets/bhmarketer/techbullion-faq-dataset
Docs	https://bhmarketer-techbullion-faq-generator.readthedocs.io
Website	https://bhmarketer.ai

License

MIT, BHMarketer.ai.