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

Investigating the citing communities around three leading health-system frameworks

Of numerous proposed frameworks for analyzing and impacting health systems, three stand out for the large number of publications that cite them and for their links to influential international institutions: Murray and Frenk (Bull World Health Organ 78:717-31, 2000) connected initially to the World Health Organization (WHO) and then to the Global Burden of Disease Project; Roberts et al. (Getting health reform right: a guide to improving performance and equity, Oxford University Press, Oxford, 2004) sponsored by the World Bank/Harvard Flagship Program; and de Savigny and Adam (Systems thinking for health systems strengthening, WHO, 2009) linked to the WHO and the Alliance for Health Policy and Systems Research. In this paper, we examine the citation communities that form around these works to better understand the underlying logic of these citation grouping as well as the dynamics of Global Health research on health systems. We conclude that these groupings are largely independent of one another, reflecting a range of factors including the goals of each framework and the problems that it was meant to explore, the prestige and authority of institutions and individuals associated with these frameworks, and the intellectual and geographic proximity of the citing researchers to each other and to the framework authors.

Cracking double-blind review: Authorship attribution with deep learning

Double-blind peer review is considered a pillar of academic research because it is perceived to ensure a fair, unbiased, and fact-centered scientific discussion. Yet, experienced researchers can often correctly guess from which research group an anonymous submission originates, biasing the peer-review process. In this work, we present a transformer-based, neural-network architecture that only uses the text content and the author names in the bibliography to attribute an anonymous manuscript to an author. To train and evaluate our method, we created the largest authorship-identification dataset to date. It leverages all research papers publicly available on arXiv amounting to over 2 million manuscripts. In arXiv-subsets with up to 2,000 different authors, our method achieves an unprecedented authorship attribution accuracy, where up to 73% of papers are attributed correctly. We present a scaling analysis to highlight the applicability of the proposed method to even larger datasets when sufficient compute capabilities are more widely available to the academic community. Furthermore, we analyze the attribution accuracy in settings where the goal is to identify all authors of an anonymous manuscript. Thanks to our method, we are not only able to predict the author of an anonymous work but we also provide empirical evidence of the key aspects that make a paper attributable. We have open-sourced the necessary tools to reproduce our experiments.

Scientometric Analysis and Classification of Research Using Convolutional Neural Networks: A Case Study in Data Science and Analytics

With the increasing development of published literature, classification methods based on bibliometric information and traditional machine learning approaches encounter performance challenges related to overly coarse classifications and low accuracy. This study presents a deep learning approach for scientometric analysis and classification of scientific literature based on convolutional neural networks (CNN). Three dimensions, namely publication features, author features, and content features, were divided into explicit and implicit features to form a set of scientometric terms through explicit feature extraction and implicit feature mapping. The weighted scientometric term vectors are fitted into a CNN model to achieve dual-label classification of literature based on research content and methods. The effectiveness of the proposed model is demonstrated using an application example from the data science and analytics literature. The empirical results show that the scientometric classification model proposed in this study performs better than comparable machine learning classification methods in terms of precision, recognition, and F1-score. It also exhibits higher accuracy than deep learning classification based solely on explicit and dominant features. This study provides a methodological guide for fine-grained classification of scientific literature and a thorough investigation of its practice.

A hybrid group-based movie recommendation framework with overlapping memberships

Recommender Systems (RS) are widely used to help people or group of people in finding their required information amid the issue of ever-growing information overload. The existing group recommender approaches consider users to be part of a single group only, but in real life a user may be associated with multiple groups having conflicting preferences. For instance, a person may have different preferences in watching movies with friends than with family. In this paper, we address this problem by proposing a Hybrid Two-phase Group Recommender Framework (HTGF) that takes into consideration the possibility of users having simultaneous membership of multiple groups. Unlike the existing group recommender systems that use traditional methods like K-Means, Pearson correlation, and cosine similarity to form groups, we use Fuzzy C-means clustering which assigns a degree of membership to each user for each group, and then Pearson similarity is used to form groups. We demonstrate the usefulness of our proposed framework using a movies data set. The experiments were conducted on MovieLens 1M dataset where we used Neural Collaborative Filtering to recommend Top-k movies to each group. The results demonstrate that our proposed framework outperforms the traditional approaches when compared in terms of group satisfaction parameters, as well as the conventional metrics of precision, recall, and F-measure.

A Novel Deep Neural Network-Based Approach to Measure Scholarly Research Dissemination Using Citations Network

We investigated the scientific research dissemination by analyzing the publications and citation data, implying that not all citations are significantly important. Therefore, as alluded to existing state-of-the-art models that employ feature-based techniques to measure the scholarly research dissemination between multiple entities, our model implements the convolutional neural network (CNN) with fastText-based pre-trained embedding vectors, utilizes only the citation context as its input to distinguish between important and non-important citations. Moreover, we speculate using focal-loss and class weight methods to address the inherited class imbalance problems in citation classification datasets. Using a dataset of 10 K annotated citation contexts, we achieved an accuracy of 90.7% along with a 90.6% f1-score, in the case of binary classification. Finally, we present a case study to measure the comprehensiveness of our deployed model on a dataset of 3100 K citations taken from the ACL Anthology Reference Corpus. We employed state-of-the-art graph visualization open-source tool Gephi to analyze the various aspects of citation network graphs, for each respective citation behavior.