Citance-based retrieval and summarization using IR and machine learning

Karimi, Samaneh; Moraes, Luis F. T.; Das, Avisha; Shakery, Azadeh; Verma, Rakesh M.

doi:10.1007/s11192-018-2785-8

Cited by 12 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These reviews have usually focused on conventional challenges, such as citation behaviour, the role of citations, and citation classification. None have focused, however, on approaches using Machine Learning (ML) and Natural Language Processing (NLP) for in-text citation analyses (Ding et al, 2013;Jeong et al, 2014;Yin et al, 2011), classification of citations (Cohen et al, 2006), citation sentiment analysis (Hernández & Gómez, 2015;Yousif et al, 2019), and citation summarisation (Gambhir & Gupta, 2017;Karimi et al, 2018).…”

Section: )mentioning

confidence: 99%

A decade of in-text citation analysis based on natural language processing and machine learning techniques: an overview of empirical studies

et al. 2021

View full text Add to dashboard Cite

show abstract

Section: )mentioning

confidence: 99%

A decade of in-text citation analysis based on natural language processing and machine learning techniques: an overview of empirical studies

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Readability metrics are an automatic and easy measurement of text difficulty [20], [27]. Readability scores like Flesch Reading Ease (fre) [22] and Dale-Chall Readability score (dcr) [32] attempt to quantify the level of difficulty of a text with respect to the reader's education level.…”

Section: ) Readability Measuresmentioning

confidence: 99%

“…For narrowing down the models and the best set of sampling hyperparameter (k and t) combinations, we look at the t-statistic and p-value given by the one-sample T-test of statistical significance. 20 The aim is to find the fine-tuned and pre-trained model(s) and (k, t) where the generated samples have a mean FRE value that is statistically similar to that of the human written references (approx. 64.7).…”

Section: A Flesch Reading Easementioning

confidence: 99%

Can Machines Tell Stories? A Comparative Study of Deep Neural Language Models and Metrics

Das

Verma

2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

Massive textual content has enabled rapid advances in natural language modeling. The use of pre-trained deep neural language models has significantly improved natural language understanding tasks. However, the extent to which these systems can be applied to content generation is unclear. While a few informal studies have claimed that these models can generate 'high quality' readable content, there is no prior study on analyzing the generated content from these models based on sampling and fine-tuning hyperparameters. We conduct an in-depth comparison of several language models for open-ended story generation from given prompts. Using a diverse set of automated metrics, we compare the performance of transformer-based generative models -OpenAI's GPT2 (pre-trained and fine-tuned) and Google's pretrained Transformer-XL and XLNet to human-written textual references. Studying inter-metric correlation along with metric ranking reveals interesting insights -the high correlation between the readability scores and word usage in the text. A study of the statistical significance and empirical evaluations between the scores (human and machine-generated) at higher sampling hyperparameter combinations (t = {0.75, 1.0}, k = {100, 150, 250}) reveal that the top pre-trained and fine-tuned models generated samples condition well on the prompt with an increased occurrence of unique and difficult words. The GPT2-medium model fine-tuned on the 1024 Byte-pair Encoding (BPE) tokenized version of the dataset along with pre-trained Transformer-XL models generated samples close to human written content on three metrics: prompt-based overlap, coherence, and variation in sentence length. A study of overall model stability and performance shows that fine-tuned GPT2 language models have the least deviation in metric scores from human performance.

show abstract

“…Clustering scientific documents aims to organise the set of documents into groups, such that documents in a single group are similar to each other in comparison to the documents in other groups (Lawrence, Bollacker, & Giles, 1999;Thijs and Glänzel, 2018). The clustering of scientific documents is crucial for several tasks, such as summarisation (Karimi et al, 2018), recommendation systems (Habib and Afzal, 2019), semantic understanding of scientific research (Shardlow et al, 2018), classification of scientific documents (Heffernan, K., & Teufel, 2018), and information retrieval systems for digital libraries (Safder & Hassan, 2019). However, the clustering of related scientific documents in growing scholar big data is a challenging task (Hassan and Haddawy, 2013;2015).…”

Section: Introductionmentioning

confidence: 99%

Tweet Coupling: a social media methodology for clustering scientific publications

et al. 2020

View full text Add to dashboard Cite

We argue that classic citation-based scientific document clustering approaches, like co-citation or bibliographic coupling, lack to leverage the social-usage of the scientific literature originate through online information dissemination platforms, such as Twitter. In this paper, we present the methodology tweet coupling, which measures the similarity between two or more scientific documents if one or more Twitter users mention them in the tweet(s). We evaluate our proposal on an altmetric dataset, which consists of 3,081 scientific documents and 8,299 unique Twitter users. By employing the clustering approaches of bibliographic coupling and tweet coupling, we find the relationship between the bibliographic and tweet coupled scientific documents. Further, using VOSviewer, we empirically show that tweet coupling appears to be a better clustering methodology to generate cohesive clusters since it groups similar documents from the subfields of the selected field, in contrast to the bibliographic coupling approach that groups cross-disciplinary documents in the same cluster.

show abstract

Citance-based retrieval and summarization using IR and machine learning

Cited by 12 publications

References 20 publications

A decade of in-text citation analysis based on natural language processing and machine learning techniques: an overview of empirical studies

A decade of in-text citation analysis based on natural language processing and machine learning techniques: an overview of empirical studies

Can Machines Tell Stories? A Comparative Study of Deep Neural Language Models and Metrics

Tweet Coupling: a social media methodology for clustering scientific publications

Contact Info

Product

Resources

About