Automatic Generation of Citation Texts in Scholarly Papers: A Pilot Study

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6181–6190 July 5 - 10, 2020. c⃝2020 Association for Computational Linguistics 6181 Automatic Generation of Citation Texts in Scholarly Papers: A Pilot Study Xinyu Xing, Xiaosheng Fan, Xiaojun Wan Wangxuan Institute of Computer Technology, Peking University The MOE Key Laboratory of Computational Linguistics, Peking University {xingxinyu,fanxiaosheng,wanxiaojun}@pku.edu.cn Abstract In this paper, we study the challenging prob- lem of automatic generation of citation texts in scholarly papers. Given the context of a cit- ing paper A and a cited paper B, the task aim- s to generate a short text to describe B in the given context of A. One big challenge for ad- dressing this task is the lack of training data. Usually, explicit citation texts are easy to ex- tract, but it is not easy to extract implicit cita- tion texts from scholarly papers. We thus first train an implicit citation text extraction mod- el based on BERT and leverage the model to construct a large training dataset for the cita- tion text generation task. Then we propose and train a multi-source pointer-generator net- work with cross attention mechanism for cita- tion text generation. Empirical evaluation re- sults on a manually labeled test dataset veri- fy the efficacy of our model. This pilot study confirms the feasibility of automatically gen- erating citation texts in scholarly papers and the technique has the great potential to help re- searchers prepare their scientific papers. 1 Introduction A scientific paper usually needs to cite a lot of reference papers and introduce each reference pa- per with some text. In this study, the text describ- ing a reference paper is called citation text. A researcher usually needs to find relevant papers he wants to cite and write some text to introduce them when writing a scientific paper. However, the process of writing citation texts is tedious and time-consuming. In order to reduce the burden of researchers, we propose and try to address the task of automatic citation text generation. Automatic generation of citation texts in schol- arly papers is a challenging and meaningful task, however, there are very few studies investigating this problem. Given a cited paper B and the con- text in a citing paper A (i.e., the sentences before and after a specific position in paper A), the task aims to generate a short text to describe B with respect to the given context in A. The task is like the task of scholarly paper summarization (Luh- n, 1958; Edmundson, 1969; Qazvinian and Radev, 2008; Mei and Zhai, 2008). Both of the two tasks aim to produce a text to describe the cited paper B. The major difference between the two tasks is that the citation texts reflect not only the salient content of B, but also the context of A. Different citing pa- pers usually have different descriptions of the same cited paper. Sometimes one paper may cite another paper several times in different positions but give different descriptions …

Content truncated to 500 words