EmHash: Hashtag Recommendation using Neural Network based on BERT Embedding

Kaviani, Mohadeseh; Rahmani, Hossein

doi:10.1109/icwr49608.2020.9122275

Cited by 18 publications

(10 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the Internet, there is an abundance of content regarding the implementation of neural networks for text translation and summarization similar to this work [Kaviani and Rahmani 2020;Li et al 2016Li et al , 2019Yang et al 2019].…”

Section: Related Workmentioning

confidence: 80%

“…Over the last 3 years, research on hashtag recommendations has become increasingly common [Li et al 2016[Li et al , 2019Yang et al 2019] [Kaviani andRahmani 2020]. Our approach differs from other works primarily because we used a corpus of Ecommerce reviews instead of content from social media, we used classical and alternative metrics to quantify the results, and we used a novel approach to generate input for the BERT model.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Text-to-hashtag Generation using Seq2Seq Learning

Camargo

Carvalho²,

Felipe

2021

Anais Do XV Brazilian E-Science Workshop (BRESCI 2021)

View full text Add to dashboard Cite

In this paper, we studied whether models based on BiLSTM and BERT can predict hashtags in Brazilian Portuguese for Ecommerce websites. Hashtags have a sizable financial impact on Ecommerce. We processed a corpus of Ecommerce reviews as inputs, and predicted hashtags as outputs. We evaluated the results using four quantitative metrics: NIST, BLEU, METEOR and a crowdsourced score. A word cloud was used as a qualitative metric. While all computer-generated metrics (NIST, BLEU and METEOR) indicated bad results, the crowdsourced results produced amazing scores. We concluded that the texts predicted by the neural networks are very promising for use as hashtags for products on Ecommerce websites. The code for this work is available at https://github.com/augustocamargo/text-to-hashtag.

show abstract

Section: Related Workmentioning

confidence: 80%

Section: Related Workmentioning

confidence: 99%

Text-to-hashtag Generation using Seq2Seq Learning

Camargo

Carvalho²,

Felipe

2021

Anais Do XV Brazilian E-Science Workshop (BRESCI 2021)

View full text Add to dashboard Cite

show abstract

“…For a query tweet, they calculated the distances between a query tweet and the centroids of the tweet clusters and extracted candidate hashtags from the closest cluster and put forward recommendations based on the hashtags' popularity values. With the increase in popularity of BERT, it has been used by Kaviani et al [40] in generating the embedding of tweets.…”

Section: Tweet Similarity Based Methodsmentioning

confidence: 99%

Hashtag Recommendation Methods for Twitter and Sina Weibo: A Review

2021

View full text Add to dashboard Cite

Hashtag recommendation suggests hashtags to users while they write microblogs in social media platforms. Although researchers have investigated various methods and factors that affect the performance of hashtag recommendations in Twitter and Sina Weibo, a systematic review of these methods is lacking. The objectives of this study are to present a comprehensive overview of research on hashtag recommendation for tweets and present insights from previous research papers. In this paper, we search for articles related to our research between 2010 and 2020 from CiteSeer, IEEE Xplore, Springer and ACM digital libraries. From the 61 articles included in this study, we notice that most of the research papers were focused on the textual content of tweets instead of other data. Furthermore, collaborative filtering methods are seldom used solely in hashtag recommendation. Taking this perspective, we present a taxonomy of hashtag recommendation based on the research methodologies that have been used. We provide a critical review of each of the classes in the taxonomy. We also discuss the challenges remaining in the field and outline future research directions in this area of study.

show abstract

“…This model is trained with cross-entropy to maximize the likelihood of ground truth tags, and produces top-k tags given the [CLS] representation. We compare two models, BR-EF and BR-LFT, for our early fusion (EF) and late fusion (LF) taken by previous models (Weston et al, 2014;Gong and Zhang, 2016;Wu et al, 2018a;Zhang et al, 2019;Yang et al, 2020a;Kaviani and Rahmani, 2020), respectively. BR-EF takes the input C (Eq.…”

Section: Baselinesmentioning

confidence: 99%

“…A conventional approach to recommendation is a ranking scheme that returns top-k relevant target items for a given query or a user profile. As such, recommending tags (e.g., hashtags and labeled tags), which is the main focus of this paper, has been treated as a ranking problem (Weston et al, 2014;Gong and Zhang, 2016;Wu et al, 2018a;Wang et al, 2019;Zhang et al, 2019;Yang et al, 2020a;Kaviani and Rahmani, 2020). These approaches, however, neglect inter-dependency among the tags (see Figure 1)in a way conventional information retrieval techniques do for ranking.…”

Section: Introductionmentioning

confidence: 99%

Leveraging Order-Free Tag Relations for Context-Aware Recommendation

Kang¹,

Kim²,

Shin³

et al. 2021

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Tag recommendation relies on either a ranking function for top-k tags or an autoregressive generation method. However, the previous methods neglect one of two seemingly conflicting yet desirable characteristics of a tag set: orderlessness and inter-dependency. While the ranking approach fails to address the interdependency among tags when they are ranked, the autoregressive approach fails to take orderlessness into account because it is designed to utilize sequential relations among tokens. We propose a sequence-oblivious generation method for tag recommendation, in which the next tag to be generated is independent of the order of the generated tags and the order of the ground truth tags occurring in training data. Empirical results on two different domains, Instagram and Stack Overflow, show that our method is significantly superior to the previous approaches.

show abstract

EmHash: Hashtag Recommendation using Neural Network based on BERT Embedding

Cited by 18 publications

References 34 publications

Text-to-hashtag Generation using Seq2Seq Learning

Text-to-hashtag Generation using Seq2Seq Learning

Hashtag Recommendation Methods for Twitter and Sina Weibo: A Review

Leveraging Order-Free Tag Relations for Context-Aware Recommendation

Contact Info

Product

Resources

About