A Web Worth of Data: Common Crawl for NLP

April 24, 2015

It's a non-profit that makes

web data

freely accessible to anyone

Each crawl archive is billions of pages:


February crawl archive is

1.9 billion web pages

~154 terabytes uncompressed

Released

totally free
without additional
intellectual property restrictions

(lives on Amazon Public Data Sets)

Common Crawl File Formats

  • WARC
    + Raw HTTP response headers
    + Raw HTTP responses

  • WAT
    + HTML head data
    + HTTP header fields
    + Extracted links / script tags

  • WET
    + Extracted text

Common Crawl WET


WET (Web Extracted Text) is released in 
the crawl archive each month

Data attempts to cover widest range of use cases


No distinction between header / navigation / content:

  • Does not remove boilerplate
  • Does not re-format text as appears in browser

Origins of Common Crawl


Common Crawl founded in 2007
by Gil Elbaz (Applied Semantics / Factual)


Google and Microsoft were the powerhouses


Goal: Democratize and simplify access to
"the web as a dataset"

Open Data and Open Source


Data powers the algorithms in our field


How can we have an even playing field for innovation without access to such data?
(Can you replicate work without the data..?)


More data can beat better algorithms
(Banko and Brill, 2001)

Common Crawl for NLP


The web is largely unannotated,
so how are people using it for NLP?


(a) Use extracted text for unsupervised algorithms

 (b) Filter it into being semi-annotated or annotated
(big data ⇒ filter ⇒ curated smaller dataset)

Examples of Previous Work


Unsupervised Algorithms

+ N-gram & language models
+ GloVe: Global Vectors for Word Representation

Filtering

+ Web tables for gazetteers
+ Dirt Cheap Web-Scale Parallel Text
+ Extracting US phone numbers

    N-gram Counts & Language Models from the Common Crawl
    Christian Buck, Kenneth Heafield, Bas van Ooyen
    (
    Edinburgh, Stanford, Owlin BV)


    Processed all the text of Common Crawl to produce 975 billion deduplicated tokens

    Google N-gram Dataset (Web 1T) consists of
    1 trillion tokens

    N-gram Counts & Language Models from the Common Crawl


    Improvement over Google N-grams (2006):
    • Inclusion of low count entries 
    • Deduplication to reduce boilerplate

    "Google has shared a deduplicated version ...
    in limited contexts, but it was never publicly released."
    -- N-gram Counts & Language Models from the Common Crawl (Buck et al.)

    N-gram Counts & Language Models from the Common Crawl



    "The advantages of structured text do not outweigh the extra computing power needed to process them."
    -- 
    N-gram Counts & Language Models from the Common Crawl (Buck et al.)

    N-gram Counts & Language Models from the Common Crawl


    English (23TB), German (1.02TB), Spanish (986GB), French (912GB), Japanese (577GB), Russian (537GB), Polish (334GB), Italian (325GB) ...

    Only 0.14% of the corpus was Finnish, yet yielded a useful corpus of 47GB.

    42 languages with >10GB

    73 languages with >1GB

    N-gram & Language Models


    Sentence level deduplication led to a removal of 80% of the English corpus, lower for other languages
    (in line with Bergsma et al. (2010))


    Before preprocessing (English): 23.62 TB


    After preprocessing (English): 5.14 TB

    (59 billion lines, 975 billion tokens)

    N-gram & Language Models

    Substantial improvement in perplexity


    N-gram & Language Models

    "...even though the web data is quite noisy even limited amounts give improvements."


    N-gram & Language Models


    Project data was released at
    http://statmt.org/ngrams

    • Raw text split by language

    • Deduped text split by language

    • Resulting language models

    GloVe: Global Vectors for Word Representation

    Jeffrey Pennington, Richard Socher, Christopher D. Manning


    Word vector representations:

    king - queen = man - woman

    king - man + woman = queen

    (produces dimensions of meaning)



    GloVe: Global Vectors for Word Representation 


    Trained on non-zero entries of a
    global word-word co-occurrence matrix 

    Populating matrix requires a single pass
    Subsequent training is far faster


    GloVe = O(|C|⁰⋅⁸)

    On-line window-based (i.e. word2vec) = O(|C|) 

    GloVe On Various Corpora

    • Semantic: "Athens is to Greece as Berlin is to _?" 
    • Syntactic: "Dance is to dancing as fly is to_?" 

    GloVe over Big Data

    GloVe using 42 billion tokens from Common Crawl outperformed word2vec w/ 100 billion tokens (Google News)


    Largest GloVe model to prove scalability uses
    840 billion tokens from Common Crawl

    Source code and pre-trained models at
    http://www-nlp.stanford.edu/projects/glove/

    Mix and Match: Word Vectors

    • More data, less fine tuning needed
    • Best model: mix of all excl. word2vec


    Examples of Previous Work


    Unsupervised Algorithms

    + N-gram & language models
    + GloVe: Global Vectors for Word Representation

    Filtering

    + Web tables for gazetteers
    + Dirt Cheap Web-Scale Parallel Text
    + Extracting US phone numbers

    Gazetteers for NER


    + Want the widest variety of topics possible

    + Aim to keep them modern / up to date

    + Capture relationships between similar words
       (disambiguation)

    Google Sets


    Web tables as a source of gazetteers + relations


    Querying ["cat"],

    returns ["dog", "bird", "horse", "rabbit", ...]

    Querying ["cat", "ls"],

    returns ["cd", "head", "cut", "vim", ...]


    Web Data Commons Web Tables


    Extracted 11.2 billion tables from WARC files,
    filtered to keep relational tables via trained classifier


    Only 1.3% of the original data was kept,
    yet it still remains hugely valuable


    Resulting dataset:
    11.2 billion tables ⇒ 147 million relational web tables

    Web Data Commons Web Tables


    Popular column headers: name, title, artist, location, model, manufacturer, country ...

    Released at webdatacommons.org/webtables/

    Web Data Commons Web Tables


    Web-Scale Parallel Text


    Dirt Cheap Web-Scale Parallel Text from the Common Crawl (Smith et al.)

    Processed all text from URLs of the style:
    website.com/[langcode]/

    "...nothing more than a set of common two-letter language codes ... [we] mined 32 terabytes ... in just under a day"

    Web-Scale Parallel Text



    (source = foreign language, target = English)

    Web-Scale Parallel Text


    Both EuroParl & United Nations are large and well curated parallel texts,

    but both have very specific domains & genres.


    Web-Scale Parallel Text


    "...resulting in improvements of up to 1.5 BLEU on standard test sets, and 5 BLEU on test sets outside of the news domain."

    Minimal cleaning & filtering still resulted in a substantial improvement in SMT performance


    Manual inspection across three languages:
    80% of the data contained good translations



    Extracting US Phone Numbers


    "Let's use Common Crawl to help match businesses from Yelp's database to the possible web pages for those businesses on the Internet."


    Yelp extracted ~748 million US phone numbers from the Common Crawl December 2014 dataset

    Regular expression over extracted text (WET files)

    Extracting US Phone Numbers


    Total complexity:
    134 lines of Python
    Total time: 1 hour (20 × c3.8xlarge)
    Total cost: $10.60 (Python using EMR)

    Matched against Yelp's database:
    • 48% had exact URL matches
    • 61% had matching domains

    More details (and full code) on Yelp's blog post:
    Analyzing the Web For the Price of a Sandwich

    WikiReverse

    Created by volunteer Ross Fairbanks for fun

    Task: Find hyperlinks to Wikipedia from the web

    Result: Dataset of over 36 million links

    Code and data released online at wikireverse.org


    Similar work by UMass and Google Research:
    Wikilinks: A Large-scale Cross-Document Coreference Corpus Labeled via Links to Wikipedia

    Common Crawl's Derived Datasets


    Natural language processing:

    Large scale web analysis:

    and a million more use cases!

    Why am I so excited..?

    Open data is catching on!

    Even playing field for academia and industry


    Common Crawl releases their dataset
    and brilliant people build on top of it

    Challenge: Parser training data

    Automatic Acquisition of Training Data for Statistical Parsers (Howlett and Curran, 2008)

    Use knowledge base of facts or simple sentences:
    "Mozart was born in 1756."

    Parse more complex sentences with dep constraints:
    "Wolfgang Amadeus Mozart (baptized Johannes Chrysostomus Wolfgangus Theophilus) was born in Salzburg in 1756, the second survivor out of six children."


    Read more at
    commoncrawl.org


    Stephen Merity
    stephen@commoncrawl.org
    commoncrawl.org

    A Web Worth of Data: Common Crawl for NLP

    By smerity

    A Web Worth of Data: Common Crawl for NLP

    • 1,296
    Loading comments...

    More from smerity