Intel-WebCorpus
Collection
Full Collection of Intel WebCorpus • 1 item • Updated
Datasets:
url stringlengths 51 411 | final_url stringlengths 53 404 | status int32 200 200 | tier stringclasses 2
values | fetch_time_ms int32 273 221k | bytes int32 2.05k 39.8M | content_hash stringlengths 64 64 | links listlengths 0 465 | title stringlengths 9 179 ⌀ | description stringlengths 0 1.11k ⌀ | og_title stringlengths 0 152 ⌀ | og_description stringlengths 0 303 ⌀ | og_image stringclasses 568
values | canonical stringlengths 32 396 ⌀ | lang stringclasses 3
values | word_count int32 0 928k | html stringlengths 1.22k 20.8M | markdown stringlengths 0 17.8M |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Pc-freezing-after-startup-Arc-B580/m-p/1754239/highlight/true#M37325 | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Pc-freezing-after-startup-Arc-B580/m-p/1754239/highlight/true | 200 | http | 4,651 | 693,388 | e591936730ab18ea9118d46c2b80330b79014d56182ad317b1d421809c9f52a6 | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | Re: Pc freezing after startup Arc B580 - Intel Community | "Hi everyone, I just want to confirm, rolling back worked for me, I had the exact same issue and I t(...TRUNCATED) | Re: Pc freezing after startup Arc B580 | "Hi everyone, I just want to confirm, rolling back worked for me, I had the exact same issue and (...TRUNCATED) | null | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Pc-freezing-after-startup-Arc-B580/m-p/1754239#M37325 | en | 18,644 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/cipcp26785/rss/message?board.id=server-products&message.id=27427 | https://community.intel.com/cipcp26785/rss/message?board.id=server-products&message.id=27427 | 200 | http | 9,812 | 6,519 | ddbb331f0336caf0de52ab86816678c9b29c0d4d40a833adf36078a880a780c9 | [] | topic Intel VROC Mixed SSD Types in Intel® Xeon® Processor and Server Products | null | null | null | null | null | null | 476 | "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<rss xmlns:content=\"http://purl.org/rss/1.0/modules/co(...TRUNCATED) | " topic Intel VROC Mixed SSD Types in Intel® Xeon® Processor and Server Products https://community(...TRUNCATED) |
https://community.intel.com/cipcp26785/rss/message?board.id=developing-games-graphics&message.id=2574 | https://community.intel.com/cipcp26785/rss/message?board.id=developing-games-graphics&message.id=2574 | 200 | http | 5,205 | 19,431 | 914caa42c6e885565213351d912b819f5d4ed63b3b0d08b535ef7984168a6bb7 | [] | "topic Re: Driver level bugs causing issues in Minecraft & OpenGL shaders in Developing Games on(...TRUNCATED) | null | null | null | null | null | null | 1,605 | "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<rss xmlns:content=\"http://purl.org/rss/1.0/modules/co(...TRUNCATED) | " topic Re: Driver level bugs causing issues in Minecraft & OpenGL shaders in Developing Games o(...TRUNCATED) |
https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Detroit-Become-Human-is-not-supported-by-ARC-B580-And-CS2-issues/m-p/1722731/highlight/true#M30286 | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Detroit-Become-Human-is-not-supported-by-ARC-B580-And-CS2-issues/m-p/1722731/highlight/true | 200 | http | 8,461 | 642,708 | 09896983a557c77d5321c3474d541bc34fdd38a2069ebcf0198a5caea565e458 | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | Re: Re:Detroit: Become Human is not supported by ARC B580. And CS2 issues - Intel Community | "Hello JeanetteC, Thank you for your patience and for guiding me through the troubleshooting process(...TRUNCATED) | Re: Re:Detroit: Become Human is not supported by ARC B580. And CS2 issues | "Hello JeanetteC, Thank you for your patience and for guiding me through the troubleshooting process(...TRUNCATED) | null | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Detroit-Become-Human-is-not-supported-by-ARC-B580-And-CS2-issues/m-p/1722731#M30286 | en | 17,522 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/t5/Intel-Confidential-Computing/SGX-test-application-missing-sgx-urts-dll/td-p/1265961?attachment-id=72275 | https://community.intel.com/t5/Intel-Confidential-Computing/SGX-test-application-missing-sgx-urts-dll/td-p/1265961?attachment-id=72275 | 200 | http | 661 | 433,658 | 4a816e696f4afa90e5738f7077bb03e1f9e6ff05bf2215ea81d3506783e01ef8 | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | Solved: SGX test application missing sgx_urts.dll - Intel Community | "Solved: First, here is my setup : My CPU is the Intel(R) Core(TM) i7-6700 CPU which is SGX capable.(...TRUNCATED) | SGX test application missing sgx_urts.dll | "First, here is my setup : My CPU is the Intel(R) Core(TM) i7-6700 CPU which is SGX capable. As I h(...TRUNCATED) | null | https://community.intel.com/t5/Intel-Confidential-Computing/SGX-test-application-missing-sgx-urts-dll/m-p/1265961#M4383 | en | 11,689 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/t5/Blogs/Products-and-Solutions/Security/IPAS-Intel-Debug-Technology/post/1332523 | https://community.intel.com/t5/Blogs/Products-and-Solutions/Security/IPAS-Intel-Debug-Technology/post/1332523 | 200 | http | 1,285 | 108,662 | 7ffb0c74ed1907cf1b5b19036ab5d76f8a9ccdea717bc6b263ced384e01606cf | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | IPAS: Intel Debug Technology - Intel Community | "Hi everyone, In a previous post , we provided some high level information about Intel debugging too(...TRUNCATED) | IPAS: Intel Debug Technology | "Hi everyone, In a previous post, we provided some high level information about Intel debugging too(...TRUNCATED) | null | https://community.intel.com/t5/Blogs/Products-and-Solutions/Security/IPAS-Intel-Debug-Technology/post/1332523 | en | 3,365 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" class=\"no-js\">\n\t<(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Monster-Hunter-Wilds-Performance-Issues-A770-16gb-on-latest/td-p/1673492 | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Monster-Hunter-Wilds-Performance-Issues-A770-16gb-on-latest/td-p/1673492 | 200 | http | 1,156 | 655,465 | acc05f985c01eb222d66fb03b5ab4ca4c65d0cdc7447a1df417b0bf91dee7365 | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | Monster Hunter Wilds - Performance Issues A770 16gb on latest drivers - Intel Community | "Hi everyone, I've been experiencing some serious performance issues while playing Monster Hunter Wi(...TRUNCATED) | Monster Hunter Wilds - Performance Issues A770 16gb on latest drivers | "Hi everyone, I've been experiencing some serious performance issues while playing Monster Hunter Wi(...TRUNCATED) | null | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/Monster-Hunter-Wilds-Performance-Issues-A770-16gb-on-latest/m-p/1673492#M22077 | en | 17,818 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/rss/message?board.id=embedded-core-processors&message.id=5991 | https://community.intel.com/rss/message?board.id=embedded-core-processors&message.id=5991 | 200 | http | 1,845 | 9,236 | 2cbfeb14d2b24307160704e5e136ae999f9e9877d991636bcf3e5c433df7e622 | [] | "topic Re: My i7 14th gen laptop hp omen 16 rtx 5060 is unable to launch games quickly like it takes(...TRUNCATED) | null | null | null | null | null | null | 627 | "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<rss xmlns:content=\"http://purl.org/rss/1.0/modules/co(...TRUNCATED) | " topic Re: My i7 14th gen laptop hp omen 16 rtx 5060 is unable to launch games quickly like it take(...TRUNCATED) |
https://community.intel.com/t5/Intel-Gaudi-AI-Accelerator/Run-the-AC-benchmark-test-According-to-the-test-log-the-NPU/m-p/1747077?attachment-id=123636 | https://community.intel.com/t5/Intel-Gaudi-AI-Accelerator/Run-the-AC-benchmark-test-According-to-the-test-log-the-NPU/m-p/1747077?attachment-id=123636 | 200 | http | 14,235 | 186,600 | 9b9579ed92957b4b8df07d5bebc9ddd4cda55d28d317b785c657acb1c9e5974a | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | "Re: Run the AC benchmark test. According to the test log, the NPU eTOPS value calculated is less th(...TRUNCATED) | can you update it | "Re: Run the AC benchmark test. According to the test log, the NPU eTOPS value calculated is less th(...TRUNCATED) | can you update it | null | https://community.intel.com/t5/Intel-Gaudi-AI-Accelerator/Run-the-AC-benchmark-test-According-to-the-test-log-the-NPU/m-p/1747077#M122 | en | 4,688 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/%C4%B05-1135G7-SUPPORT-XeSS/m-p/1380749#M44 | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/%C4%B05-1135G7-SUPPORT-XeSS/m-p/1380749 | 200 | http | 12,922 | 184,025 | dd304e33e64d804382f6bf7e80a1fd0a31cfa81d1776d2e5db8e7358e7157bb8 | ["https://community.intel.com/","https://community.intel.com/t5/About/bd-p/about-communities","https(...TRUNCATED) | Solved: Re: İ5-1135G7 SUPPORT XeSS - Intel Community | One post is enough. Stay with your other post: | Re: İ5-1135G7 SUPPORT XeSS | "One post is enough. Stay with your other post: https://community.intel.com/t5/Gaming-on-Intel-P(...TRUNCATED) | null | https://community.intel.com/t5/Intel-Arc-Discrete-Graphics/%C4%B05-1135G7-SUPPORT-XeSS/m-p/1380749#M44 | en | 4,593 | "<!DOCTYPE html><html prefix=\"og: http://ogp.me/ns#\" dir=\"ltr\" lang=\"en\" itemtype=\"http://sch(...TRUNCATED) | " Browse\n\n* [Support Community](/)\n\n* [About](/t5/About/bd-p/about-communities)\n* [Developer So(...TRUNCATED) |
End of preview. Expand in Data Studio
🌐 Intel WebCorpus v1
A massive, uncleaned, raw dataset crawled from Intel blogs and web properties.
Intel WebCorpus v1 provides ~4.33 GB of parsed content spanning Intel's corporate blogs, technical articles, and public communications. Notably, this dataset is intentionally uncleaned. It retains all HTML, CSS, JavaScript, boilerplates, and duplicates exactly as crawled, offering an authentic, unfiltered look at enterprise tech properties for robust pre-training and specialized parsing tasks.
📑 Table of Contents
- Dataset Summary
- How to Use
- Dataset Structure
- Data Pipeline & Architecture
- Limitations & Biases
- License
📊 Dataset Summary
Intel WebCorpus v1 was built to provide a high-quality slice of hardware, semiconductor, and corporate communications text. We systematically crawled Intel domains and blog subdirectories, resolving redirects, capturing metadata (OpenGraph), and extracting textual content.
| Total Files | 94 .parquet chunks |
| Total Size | ~4.33 GB (compressed) |
| Primary Language | English |
| Format | Apache Parquet |
| Total Tokens | ~853 Million (cl100k_base) |
| Fields Provided | 18 distinct metadata and content fields |
| Cleaning Status | Uncleaned (Contains all boilerplate, HTML, CSS, JS, and duplicates) |
🚀 How to Use
To use this dataset in your project via the datasets library, it's as simple as:
from datasets import load_dataset
# Load the entire corpus
dataset = load_dataset("sphita/Intel-WebCorpus-v1", split="train")
# Print the first document's markdown content
print(dataset[0]['markdown'])
🏗️ Dataset Structure
Schema Overview
Each row represents a uniquely crawled webpage. The corpus provides extensive metadata alongside the extracted text to allow for advanced filtering (e.g., by word count, language, or HTTP status).
| Field | Type | Description |
|---|---|---|
url |
string |
The initially requested URL. |
final_url |
string |
The resolved URL after any HTTP redirects. |
status |
int32 |
The HTTP response status code (e.g., 200). |
tier |
string |
Crawl tier or depth categorization. |
fetch_time_ms |
int32 |
Time taken to fetch the page in milliseconds. |
bytes |
int32 |
Raw size of the HTML document. |
content_hash |
string |
Hash of the content for deduplication tracking. |
links |
sequence |
List of outbound links found on the page. |
title |
string |
HTML <title> of the page. |
description |
string |
HTML <meta name="description"> content. |
og_title |
string |
OpenGraph Title (og:title). |
og_description |
string |
OpenGraph Description (og:description). |
og_image |
string |
Primary OpenGraph Image URL (og:image). |
canonical |
string |
Canonical link URL defined in the page head. |
lang |
string |
Detected or defined language (e.g., en). |
word_count |
int32 |
Number of words in the extracted markdown. |
html |
string |
The raw, uncleaned HTML body of the page. |
markdown |
string |
The uncleaned Markdown text, retaining duplicates, boilerplates, HTML, CSS, and embedded JS exactly as found. |
🔍 Click to view an example JSON record
{
"url": "https://blogs.intel.com/technology/example-post",
"final_url": "https://blogs.intel.com/technology/example-post",
"status": 200,
"tier": "Tier 1",
"fetch_time_ms": 145,
"bytes": 102450,
"content_hash": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"links": ["https://intel.com/news", "https://blogs.intel.com/contact"],
"title": "Advancing Silicon Innovation",
"description": "Intel's latest breakthroughs in semiconductor manufacturing.",
"og_title": "Advancing Silicon Innovation - Intel Blogs",
"og_description": "A deep dive into the new fab architecture.",
"og_image": "https://blogs.intel.com/images/silicon-hero.jpg",
"canonical": "https://blogs.intel.com/technology/example-post",
"lang": "en",
"word_count": 845,
"html": "<!DOCTYPE html><html><head><title>Advancing Silicon Innovation</title>...</head><body>...",
"markdown": "# Advancing Silicon Innovation\n\n<style>body{margin:0;}</style>\n\nToday, Intel announces..."
}
⚙️ Data Pipeline & Architecture
The corpus was gathered using our custom Queen Crawler (Rust), a highly concurrent, tiered web crawler built for robustness and high throughput.
🌐 1. Seed URLs ↓ 🕷️ 2. Queen Crawler (Rust) ↓ ⚖️ 3. Content Filter (Excludes Non-Text/Errors) ↓ 📄 4. Raw HTML Parsing ↓ 📝 5. Uncleaned Markdown Conversion ↓ 🗄️ 6. Parquet Storage
- Discovery: Starting from a list of root Intel domains and blog directories.
- Extraction: The raw HTML was intentionally kept completely uncleaned of duplicates, navigation, footers, or any boilerplate.
- Transformation: The resulting DOM is transformed into Markdown, explicitly retaining embedded
<script>,<style>, HTML, and CSS information without any stripping. - Storage: Output is sharded into
~50MBParquet chunks to balance I/O efficiency and dataset modularity.
⚠️ Limitations & Biases
- Temporal Scope: The dataset reflects a snapshot in time. Fast-moving tech specs or corporate statements may become outdated.
- Corporate Voice: The text is heavily skewed toward Intel's marketing, PR, and internal technical blog style. It is not representative of casual conversational text.
- Uncleaned Nature: Because the dataset retains all boilerplates, duplicates, scripts, and CSS, it requires rigorous downstream cleaning for most standard NLP tasks.
📄 License
This dataset is curated and structured for research purposes. The dataset format and metadata curation are released under the GPL-3.0 License. The underlying text, logos, and intellectual property remain the copyright of Intel Corporation.
Dataset Card maintained by @sphita
- Downloads last month
- 79