We used neural networks to detect clickbaits: You won’t believe what happened next!

Ankesh Anand; Tanmoy Chakraborty; Noseong Park

doi:10.1007/978-3-319-56608-5_46

We used neural networks to detect clickbaits: You won’t believe what happened next!

Ankesh Anand, Tanmoy Chakraborty, Noseong Park

Graduate School

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

47 Citations (Scopus)

Abstract

Online content publishers often use catchy headlines for their articles in order to attract users to their websites. These headlines, popularly known as clickbaits, exploit a user’s curiosity gap and lure them to click on links that often disappoint them. Existing methods for automatically detecting clickbaits rely on heavy feature engineering and domain knowledge. Here, we introduce a neural network architecture based on Recurrent Neural Networks for detecting clickbaits. Our model relies on distributed word representations learned from a large unannotated corpora, and character embeddings learned via Convolutional Neural Networks. Experimental results on a dataset of news headlines show that our model outperforms existing techniques for clickbait detection with an accuracy of 0.98 with F1-score of 0.98 and ROC-AUC of 0.99.

Original language	English
Title of host publication	Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings
Editors	Claudia Hauff, Joemon M. Jose, Dyaa Albakour, Ismail Sengor Altingovde, John Tait, Dawei Song, Stuart Watt
Publisher	Springer Verlag
Pages	541-547
Number of pages	7
ISBN (Print)	9783319566078
DOIs	https://doi.org/10.1007/978-3-319-56608-5_46
Publication status	Published - 2017
Event	39th European Conference on Information Retrieval, ECIR 2017 - Aberdeen, United Kingdom Duration: 2017 Apr 8 → 2017 Apr 13

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10193 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	39th European Conference on Information Retrieval, ECIR 2017
Country/Territory	United Kingdom
City	Aberdeen
Period	17/4/8 → 17/4/13

Bibliographical note

Publisher Copyright:
© Springer International Publishing AG 2017.

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-319-56608-5_46

Cite this

Anand, A., Chakraborty, T., & Park, N. (2017). We used neural networks to detect clickbaits: You won’t believe what happened next! In C. Hauff, J. M. Jose, D. Albakour, I. S. Altingovde, J. Tait, D. Song, & S. Watt (Eds.), Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings (pp. 541-547). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10193 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-56608-5_46

Anand, Ankesh ; Chakraborty, Tanmoy ; Park, Noseong. / We used neural networks to detect clickbaits : You won’t believe what happened next!. Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings. editor / Claudia Hauff ; Joemon M. Jose ; Dyaa Albakour ; Ismail Sengor Altingovde ; John Tait ; Dawei Song ; Stuart Watt. Springer Verlag, 2017. pp. 541-547 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4a77a6a97f1d4ba299f6dffe9024305d,

title = "We used neural networks to detect clickbaits: You won{\textquoteright}t believe what happened next!",

abstract = "Online content publishers often use catchy headlines for their articles in order to attract users to their websites. These headlines, popularly known as clickbaits, exploit a user{\textquoteright}s curiosity gap and lure them to click on links that often disappoint them. Existing methods for automatically detecting clickbaits rely on heavy feature engineering and domain knowledge. Here, we introduce a neural network architecture based on Recurrent Neural Networks for detecting clickbaits. Our model relies on distributed word representations learned from a large unannotated corpora, and character embeddings learned via Convolutional Neural Networks. Experimental results on a dataset of news headlines show that our model outperforms existing techniques for clickbait detection with an accuracy of 0.98 with F1-score of 0.98 and ROC-AUC of 0.99.",

author = "Ankesh Anand and Tanmoy Chakraborty and Noseong Park",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 39th European Conference on Information Retrieval, ECIR 2017 ; Conference date: 08-04-2017 Through 13-04-2017",

year = "2017",

doi = "10.1007/978-3-319-56608-5_46",

language = "English",

isbn = "9783319566078",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "541--547",

editor = "Claudia Hauff and Jose, {Joemon M.} and Dyaa Albakour and Altingovde, {Ismail Sengor} and John Tait and Dawei Song and Stuart Watt",

booktitle = "Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings",

address = "Germany",

}

Anand, A, Chakraborty, T & Park, N 2017, We used neural networks to detect clickbaits: You won’t believe what happened next! in C Hauff, JM Jose, D Albakour, IS Altingovde, J Tait, D Song & S Watt (eds), Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10193 LNCS, Springer Verlag, pp. 541-547, 39th European Conference on Information Retrieval, ECIR 2017, Aberdeen, United Kingdom, 17/4/8. https://doi.org/10.1007/978-3-319-56608-5_46

We used neural networks to detect clickbaits: You won’t believe what happened next! / Anand, Ankesh; Chakraborty, Tanmoy; Park, Noseong.
Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings. ed. / Claudia Hauff; Joemon M. Jose; Dyaa Albakour; Ismail Sengor Altingovde; John Tait; Dawei Song; Stuart Watt. Springer Verlag, 2017. p. 541-547 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10193 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - We used neural networks to detect clickbaits

T2 - 39th European Conference on Information Retrieval, ECIR 2017

AU - Anand, Ankesh

AU - Chakraborty, Tanmoy

AU - Park, Noseong

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Online content publishers often use catchy headlines for their articles in order to attract users to their websites. These headlines, popularly known as clickbaits, exploit a user’s curiosity gap and lure them to click on links that often disappoint them. Existing methods for automatically detecting clickbaits rely on heavy feature engineering and domain knowledge. Here, we introduce a neural network architecture based on Recurrent Neural Networks for detecting clickbaits. Our model relies on distributed word representations learned from a large unannotated corpora, and character embeddings learned via Convolutional Neural Networks. Experimental results on a dataset of news headlines show that our model outperforms existing techniques for clickbait detection with an accuracy of 0.98 with F1-score of 0.98 and ROC-AUC of 0.99.

AB - Online content publishers often use catchy headlines for their articles in order to attract users to their websites. These headlines, popularly known as clickbaits, exploit a user’s curiosity gap and lure them to click on links that often disappoint them. Existing methods for automatically detecting clickbaits rely on heavy feature engineering and domain knowledge. Here, we introduce a neural network architecture based on Recurrent Neural Networks for detecting clickbaits. Our model relies on distributed word representations learned from a large unannotated corpora, and character embeddings learned via Convolutional Neural Networks. Experimental results on a dataset of news headlines show that our model outperforms existing techniques for clickbait detection with an accuracy of 0.98 with F1-score of 0.98 and ROC-AUC of 0.99.

UR - http://www.scopus.com/inward/record.url?scp=85018698975&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018698975&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-56608-5_46

DO - 10.1007/978-3-319-56608-5_46

M3 - Conference contribution

AN - SCOPUS:85018698975

SN - 9783319566078

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 541

EP - 547

BT - Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings

A2 - Hauff, Claudia

A2 - Jose, Joemon M.

A2 - Albakour, Dyaa

A2 - Altingovde, Ismail Sengor

A2 - Tait, John

A2 - Song, Dawei

A2 - Watt, Stuart

PB - Springer Verlag

Y2 - 8 April 2017 through 13 April 2017

ER -

Anand A, Chakraborty T, Park N. We used neural networks to detect clickbaits: You won’t believe what happened next! In Hauff C, Jose JM, Albakour D, Altingovde IS, Tait J, Song D, Watt S, editors, Advances in Information Retrieval - 39th European Conference on IR Research, ECIR 2017, Proceedings. Springer Verlag. 2017. p. 541-547. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-56608-5_46

We used neural networks to detect clickbaits: You won’t believe what happened next!

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this