PREDIKCIJA POPULARNOSTI OBJAVA NA SAJTU 9GAG NA OSNOVU SLIKE
Ključne reči:
istraživanje podataka, analiza podataka, računarski vid, mašinsko učenje, regresiona analiza
Apstrakt
U radu je eksperimentisano sa više modela mašinskog učenja za predikciju popularnosti objave na 9gag društvenoj mreži. Fokus je na predviđanju popularnosti objava na osnovu analize slike. Slike su analizirane izvlačenjem tri grupe obeležja, koje predstavljaju: (1) skup objekata prepoznatih na slici, (2) postojanje prepoznatog popularnog šablona na slici i (3) dužinu tekstualnog sadržaja na slici. Drugi pristup analizi slika je end-to-end pristup, koji se bazira na dubokom učenju. Modeli predstavljeni u radu su deo šireg sistema koji predviđa popularnost objave na osnovu kombinovanih informacija ekstrahovanih iz slike, teksta, i metapodataka. U radu je eksperimentisano i sa više pristupa kombinovanja ovih informacija.
Reference
[1] Castaño Díaz, Carlos Mauricio. "Defining and characterizing the concept of Internet Meme." CES Psicología 6.2 (2013): 82-104.
[2] Kempe, David, Jon Kleinberg, and Éva Tardos. "Maximizing the spread of influence through a social network." Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining. 2003.
[3] https://www.crunchbase.com/organization/9gag (posećeno septembra 2020.)
[4] Meghawat, Mayank, et al. "A multimodal approach to predict social media popularity." 2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR). IEEE, 2018.
[5] Mazloom, Masoud, et al. "Multimodal popularity prediction of brand-related social media posts." Proceedings of the 24th ACM international conference on Multimedia. 2016.
[6] Borth, Damian, et al. "Sentibank: large-scale ontology and classifiers for detecting sentiment and emotions in visual content." Proceedings of the 21st ACM international conference on Multimedia. 2013.
[7] Thelwall, Mike, et al. "Sentiment strength detection in short informal text." Journal of the American society for information science and technology 61.12 (2010): 2544-2558.
[8] Hu, Jiani, Toshihiko Yamasaki, and Kiyoharu Aizawa. "Multimodal learning for image popularity prediction on social media." 2016 IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW). IEEE, 2016.
[9] Thomee, Bart, et al. "YFCC100M: The new data in multimedia research." Communications of the ACM 59.2 (2016): 64-73.
[10] Khosla, Aditya, Atish Das Sarma, and Raffay Hamid. "What makes an image popular?." Proceedings of the 23rd international conference on World wide web. 2014.
[11] Sandler, Mark, et al. "Mobilenetv2: Inverted residuals and linear bottlenecks." Proceedings of the IEEE conference on computer vision and pattern recognition. 2018.
[12] Kuznetsova, Alina, et al. "The open images dataset v4: Unified image classification, object detection, and visual relationship detection at scale." arXiv preprint arXiv:1811.00982 (2018).
[13] Simonyan, Karen, and Andrew Zisserman. "Very deep convolutional networks for large-scale image recognition." arXiv preprint arXiv:1409.1556 (2014).
[14] Deng, Jia, et al. "Imagenet: A large-scale hierarchical image database." 2009 IEEE conference on computer vision and pattern recognition. Ieee, 2009.
[15] Lee, Kevan. "The proven ideal length of every tweet, Facebook post, and headline online." Fast Company, Apr (2014).
[16]https://www.analyticsvidhya.com/blog/2018/06/comprehensive-guide-for-ensemble-models/ (posećeno septembra 2020.)
[2] Kempe, David, Jon Kleinberg, and Éva Tardos. "Maximizing the spread of influence through a social network." Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining. 2003.
[3] https://www.crunchbase.com/organization/9gag (posećeno septembra 2020.)
[4] Meghawat, Mayank, et al. "A multimodal approach to predict social media popularity." 2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR). IEEE, 2018.
[5] Mazloom, Masoud, et al. "Multimodal popularity prediction of brand-related social media posts." Proceedings of the 24th ACM international conference on Multimedia. 2016.
[6] Borth, Damian, et al. "Sentibank: large-scale ontology and classifiers for detecting sentiment and emotions in visual content." Proceedings of the 21st ACM international conference on Multimedia. 2013.
[7] Thelwall, Mike, et al. "Sentiment strength detection in short informal text." Journal of the American society for information science and technology 61.12 (2010): 2544-2558.
[8] Hu, Jiani, Toshihiko Yamasaki, and Kiyoharu Aizawa. "Multimodal learning for image popularity prediction on social media." 2016 IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW). IEEE, 2016.
[9] Thomee, Bart, et al. "YFCC100M: The new data in multimedia research." Communications of the ACM 59.2 (2016): 64-73.
[10] Khosla, Aditya, Atish Das Sarma, and Raffay Hamid. "What makes an image popular?." Proceedings of the 23rd international conference on World wide web. 2014.
[11] Sandler, Mark, et al. "Mobilenetv2: Inverted residuals and linear bottlenecks." Proceedings of the IEEE conference on computer vision and pattern recognition. 2018.
[12] Kuznetsova, Alina, et al. "The open images dataset v4: Unified image classification, object detection, and visual relationship detection at scale." arXiv preprint arXiv:1811.00982 (2018).
[13] Simonyan, Karen, and Andrew Zisserman. "Very deep convolutional networks for large-scale image recognition." arXiv preprint arXiv:1409.1556 (2014).
[14] Deng, Jia, et al. "Imagenet: A large-scale hierarchical image database." 2009 IEEE conference on computer vision and pattern recognition. Ieee, 2009.
[15] Lee, Kevan. "The proven ideal length of every tweet, Facebook post, and headline online." Fast Company, Apr (2014).
[16]https://www.analyticsvidhya.com/blog/2018/06/comprehensive-guide-for-ensemble-models/ (posećeno septembra 2020.)
Objavljeno
2020-12-25
Sekcija
Elektrotehničko i računarsko inženjerstvo
