Many useful tasks on scientific documents, such as topic classification and
citation prediction, involve corpora that span multiple scientific domains.
Typically, such tasks are accomplished by representing the text with a vector
embedding obtained from a Transformer's single CLS token. In this paper, we
argue that using multiple CLS tokens could make a Transformer better specialize
to multiple scientific domains. We present Multi2SPE: it encourages each of
multiple CLS tokens to learn diverse ways of aggregating token embeddings, then
sums them up together to create a single vector representation. We also propose
our new multi-domain benchmark, Multi-SciDocs, to test scientific paper vector
encoders under multi-domain settings. We show that Multi2SPE reduces error by
up to 25 percent in multi-domain citation prediction, while requiring only a
negligible amount of computation in addition to one BERT forward pass.

使用多个 CLS 标记对科学文档进行多领域特化，提出 Multi2SPE 方法，并在多领域下的科学论文向量编码中展示 Multi2SPE 在减少错误率上的效果。

利用多个 CLS 令牌集成编码多领域科学论文

Encoding Multi-Domain Scientific Papers by Ensembling Multiple CLS  Tokens

Predicting the number of citations of scholarly documents is an upcoming task
in scholarly document processing. Besides the intrinsic merit of this
information, it also has a wider use as an imperfect proxy for quality which
has the advantage of being cheaply available for large volumes of scholarly
documents. Previous work has dealt with number of citations prediction with
relatively small training data sets, or larger datasets but with short,
incomplete input text. In this work we leverage the open access ACL Anthology
collection in combination with the Semantic Scholar bibliometric database to
create a large corpus of scholarly documents with associated citation
information and we propose a new citation prediction model called SChuBERT. In
our experiments we compare SChuBERT with several state-of-the-art citation
prediction models and show that it outperforms previous methods by a large
margin. We also show the merit of using more training data and longer input for
number of citations prediction.

本文介绍了一种新的引用预测模型 SChuBERT，该模型利用 ACL 文献库和 Semantic Scholar 文献数据库的大型语料库进行训练，并通过实验证明，在预测引用数量方面，它比现有方法表现更优秀。

SChuBERT: 利用 BERT 编码增强引文计数预测的学术文献块

SChuBERT: Scholarly Document Chunks with BERT-encoding boost Citation  Count Prediction

Citation prediction of scholarly papers is of great significance in guiding
funding allocations, recruitment decisions, and rewards. However, little is
known about how citation patterns evolve over time. By exploring the inherent
involution property in scholarly paper citation, we introduce the Paper
Potential Index (PPI) model based on four factors: inherent quality of
scholarly paper, scholarly paper impact decaying over time, early citations,
and early citers' impact. In addition, by analyzing factors that drive citation
growth, we propose a multi-feature model for impact prediction. Experimental
results demonstrate that the two models improve the accuracy in predicting
scholarly paper citations. Compared to the multi-feature model, the PPI model
yields superior predictive performance in terms of range-normalized RMSE. The
PPI model better interprets the changes in citation, without the need to adjust
parameters. Compared to the PPI model, the multi-feature model performs better
prediction in terms of Mean Absolute Percentage Error and Accuracy; however,
their predictive performance is more dependent on the parameter adjustment.

研究了学术论文引用格局随时间演变的特点，提出了基于 “学术论文固有质量”、“论文引用的时效性”、“早期引用” 和 “早期引用的影响力” 四个因素的学术论文潜力指数模型 (PPI)，并通过分析影响论文引用增长的因素，提出了多特征的影响力预测模型，实验证明这两种模型的预测准确性得到了改善。其中，相对于多特征模型，PPI 模型的预测性能更好，且不需要调整参数，因此更好的解释了引用的变化。而多特征模型在平均绝对百分误差和准确性方面表现更好，但它们的预测效果更依赖于参数调整。

学术论文引用预测

Predicting the Citations of Scholarly Paper

The use of citation counts to assess the impact of research articles is well
established. However, the citation impact of an article can only be measured
several years after it has been published. As research articles are
increasingly accessed through the Web, the number of times an article is
downloaded can be instantly recorded and counted. One would expect the number
of times an article is read to be related both to the number of times it is
cited and to how old the article is. This paper analyses how short-term Web
usage impact predicts medium-term citation impact. The physics e-print archive
(arXiv.org) is used to test this.

本研究分析了短期网络使用情况对中期引证影响的预测作用，以文章引用计数、Web 阅读次数及 e-print 论文库为研究对象。