Get best of the week

Web2Text: deep structured extraction of web page content

Hello, Habr! I present to you the translation of the article "Web2Text: Deep Structured Boilerplate Removal" by a team of authors Thijs Vogels, Octavian-Eugen Ganea and Carsten Eickhof.


Web pages are a valuable source of information for many natural language processing and information retrieval tasks. Effectively extracting core content from these documents is critical to the performance of derived applications. To solve this problem, we introduce a new model that classifies and labels text blocks on a page HTMLas template blocks, or blocks containing main content. Our method uses the Hidden Markov model on top of potentials obtained from the features of the object model of the HTMLdocument ( Document Object Model, DOM) using convolutional neural networks ( Convolutional Neural Network, CNN). The proposed method qualitatively improves performance for extracting text data from web pages.


1. Introduction


Modern methods of natural language processing and information retrieval are highly dependent on large collections of text. The World Wide Web is an inexhaustible source of content for such applications. However, a common problem is that web pages include not only the main content (text), but also ads, hyperlink lists, navigation, previews of other articles, banners, etc. This template content often has a negative impact on the performance of a derived application [15,24]. The task of separating the main text on a web page from the rest (template) content in the literature is known as “deleting a standard template”, “segmenting a web page”, or “extracting content”. Known popular methods for this problem use rule-based algorithms or machine learning.The most successful approaches first split the input web page into text blocks, and then binary{1, 0}Labeling each block as the main content or template. In this article, we propose the Hidden Markov Model on top of neural potentials for the task of removing patterns. We use the ability of convolutional neural networks to study unary and pair potentials in blocks based on complex nonlinear combinations of signs based on DOM. During forecasting, we find the most probable block label {1, 0}, maximizing the joint probability of the sequence of labels using the Viterbi algorithm [23]. The effectiveness of our method is demonstrated on standard sets of comparative data.


. 2 . 3 , . 4 -.


2.


HTML- [7] Body Text Extractor (BTE). BTE , , HTML- -. , BTE , . , : (1) HTML, , , (2) , -.


DOM, HTML [11,19,6]. , , <table>, .


DOM . . [24] [22]. , -, -, -. .


. [10] , . HTML , , , . , , (), , (). DOM [4,21]. . [3] DOM, , . . [21] / , DOM .


«», . FIASCO . [2] (SVM) HTML- , DOM , . . [17] SVM . . [20] , , . , . CleanEval [1].


p1


. 1. Web2Text. DOM (Collapsed DOM) - , . DOM. , , : . . , , , .


, DOM. , , . , - , .


3.


— - (- ) [1]. . 1.


3.1.


, - (X) HTML-. ( DOM) Jsoup [12].


p2


. 2. DOM. : HTML, — DOM, — DOM.


DOM, i) , , ii) , , : , <br>, <checkbox>, <head>, <hr>, <iframe>, <img>, <input>. DOM. DOM- . 2 DOM, (<ul>), DOM. (, « »), . Collapsed () DOM (CDOM).


3.2.


. - , , . - , : i) HTML, ii) DOM, iii) DOM . DOM, . , , HTML. , DOM- ( #text) . , , . , Web2Text , , — .


3.3.


— , , , . , CDOM . .


. , CDOM, , CDOM. 128 , , « - <p>», « », « », « », « - » .. , , .


. 25 . . , , , 2, 3, 4 > 4. , HTML-, ..


3.4. (Convolutional Neural Network, CNN)


, , . , . pi (li = 1), pi (li = 0) , li i , . . pi, i + 1 (li = 1, li + 1 = 1), pi, i + 1 (li = 1, li + 1 = 0), pi, i + 1 (li = 0, li + 1 = 1) pi, i + 1 (li = 0, li + 1 = 0) — . .


CNN 5 , ReLU , (50, 50, 50, 10, 2) (50, 50, 50, 10, 4) . 1 (1, 1, 3, 3, 3) . CNN , , , . CNN , , , . , , . 2 , softmax. 4 , . , i . (dropout) 0,2 L2 10-4.


-:


f1


l∗i — i, θunary — , n — .


-:


f2


θpairwise — .


3.5.


- . (b0, b1, ..., bn) (l0, l1, ..., ln) ∈ {0, 1}n :


f3


λ — . λ = 0,1 . [23], CNN.


4.


. Web2Text - . , . Web2Text .


4.1.


CleanEval 2007 [1] . 188 -. (60 ) (676 ). (55 ) (5 ). 10000 , , . CleanEval : (531 ), (58 ) (148 ).


. , ( CleanEval) . “- — ” ( ). , , . (, [20]) . (, ) (-, ). .


-, 10 . - ( ). , - , « ». -. , , , , 2/3 .


4.2.


[14] 10–3 5000 . - 128 - 9 . , . , .


4.3.


Web2Text , . BTE [7] Unfluff [8] . [17,16] — , , (. 1). CRF [20] CleanEval. (Conditional Random Field, CRF) , . , 4.1, CRF - . , , , , CleanEval. CleanEval, , .


. CRF [20] 9 705 . , CNN 17 960 , CNN 12 870 . 30 830. , .


4.4.


1 . , . , Web2Text (Accuracy), Recall F1 , CleanEval. , , , 3.2. , , Web2Text CNN, .


t1


1. - CleanEval. : (55 — , 5 — , 676 — ) (531 — , 58 — , 148 — ). , .


. Web2Text 54 -; 35 DOM , 19 . Macbook Intel Core i5 2,8 .


4.5.


, , , . HTML, .


- ClueWeb12. . CW12-A 733M - (27,3 ) CW12-B 52M (1,95 ). Indri. 50 TREC Web Track 2013 [5].


t2


2. . (*) HTML. (†) , .


2 , -. HTML . , , †. , , CW12-A, , , CW12-B. - . , (QL) , (RM). , . , (BTE, article-ext, large-ext, Unfluff) , . (CRF, Web2Text) . , Web2Text 0,05. , Web2Text CleanEval, 4.1.


5.


Web2Text -. , CRF [9], , DOM . CleanEval . , , , .


6.


, - .
, .


  1. Marco Baroni, Francis Chantree, Adam Kilgarriff, and Serge Sharoff. CleanEval: a competition for cleaning web pages. In LREC, 2008.
  2. Daniel Bauer, Judith Degen, Xiaoye Deng, Priska Herger, Jan Gasthaus, Eugenie Giesbrecht, Lina Jansen, Christin Kalina, Thorben Kräger, Robert Märtin, Martin Schmidt, Simon Scholler, Johannes Steger, Egon Stemle, and Stefan Evert. FIASCO: Filtering the internet by automatic subtree classification, osnabruck. In Building and Exploring Web Corpora: Proceedings of the 3rd Web as Corpus Workshop, incorporating CleanEval, volume 4, pages 111–121, 2007.
  3. Deepayan Chakrabarti, Ravi Kumar, and Kunal Punera. Page-level template detection via isotonic smoothing. In Proceedings of the 16th international conference on World Wide Web, pages 61–70. ACM, 2007.
  4. Deepayan Chakrabarti, Ravi Kumar, and Kunal Punera. A graph-theoretic approach to webpage segmentation. In Proceedings of the 17th international conference on World Wide Web, pages 377–386. ACM, 2008.
  5. Kevyn Collins-Thompson, Paul Bennett, Fernando Diaz, Charlie Clarke, and Ellen Voorhees. Overview of the TREC 2013 web track. In Proceedings of the 22nd Text Retrieval Conference (TREC’13), 2013.
  6. Sandip Debnath, Prasenjit Mitra, Nirmal Pal, and C Lee Giles. Automatic identification of informative sections of web pages. IEEE transactions on knowledge and data engineering, 17(9):1233–1246, 2005.
  7. Aidan Finn, Nicholas Kushmerick, and Barry Smyth. Fact or fiction: Content classification for digital libraries. Unrefereed, 2001.
  8. Adam Geitgey. Unfluff – an automatic web page content extractor for node.js!, 2014.
  9. John Gibson, Ben Wellner, and Susan Lubar. Adaptive web-page content identification. In Proceedings of the 9th annual ACM international workshop on Web information and data management, pages 105–112. ACM, 2007.
  10. Thomas Gottron. Content code blurring: A new approach to content extraction. In Database and Expert Systems Application, 2008. DEXA’08. 19th International Workshop on, pages 29–33. IEEE, 2008.
  11. Suhit Gupta, Gail Kaiser, David Neistadt, and Peter Grimm. DOM-based content extraction of HTML documents. In Proceedings of the 12th international conference on World Wide Web, pages 207–214. ACM, 2003.
  12. Jonathan Hedley. Jsoup HTML parser, 2009.
  13. Rong Jin, Alex G Hauptmann, and ChengXiang Zhai. Language model for information retrieval. In Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval, pages 42–48. ACM, 2002.
  14. Diederik Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.
  15. Christian Kohlschütter. A densitometric analysis of web template content. In Proceedings of the 18th international conference on World wide web, pages 1165– 1166. ACM, 2009.
  16. Christian Kohlschütter et al. Boilerpipe – boilerplate removal and fulltext extraction from HTML pages. Google Code, 2010.
  17. Christian Kohlschütter, Peter Fankhauser, and Wolfgang Nejdl. Boilerplate detection using shallow text features. In Proceedings of the third ACM international conference on Web search and data mining, pages 441–450. ACM, 2010.
  18. Victor Lavrenko and W Bruce Croft. Relevance based language models. In Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval, pages 120–127. ACM, 2001.
  19. Shian-Hua Lin and Jan-Ming Ho. Discovering informative content blocks from web documents. In Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 588–593. ACM, 2002.
  20. Miroslav Spousta, Michal Marek, and Pavel Pecina. Victor: the web-page cleaning tool. In 4th Web as Corpus Workshop (WAC4)-Can we beat Google, pages 12–17, 2008.
  21. Fei Sun, Dandan Song, and Lejian Liao. Dom based content extraction via text density. In Proceedings of the 34th international ACM SIGIR conference on Research and development in Information Retrieval, pages 245–254. ACM, 2011.
  22. Karane Vieira, Altigran S Da Silva, Nick Pinto, Edleno S De Moura, Joao Cavalcanti, and Juliana Freire. A fast and robust method for web page template detection and removal. In Proceedings of the 15th ACM international conference on Information and knowledge management, pages 258–267. ACM, 2006.
  23. Andrew J Viterbi. Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. In The Foundations Of The Digital Wireless World: Selected Works of AJ Viterbi, pages 41–50. World Scientific, 2010.
  24. Lan Yi, Bing Liu, and Xiaoli Li. Eliminating noisy information in web pages for data mining. In Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 296–305. ACM, 2003.

More articles:


All Articles