Rationale classification for educational trading platforms

    title = "Rationale Classification for Educational Trading Platforms",
    author = {Ying, Annie  and  Duboue, Pablo},
    booktitle = "Proceedings of the First Workshop on Financial Technology and Natural Language Processing",
    month = {August},
    year = "2019",
    address = "Macao, China",
    url = "https://www.aclweb.org/anthology/W19-5503",
    pages = "14--20"

This joint work with Dr. Ying was presented at the FinNLP Workshop in conjunction with IJCAI-2019 in Macao, China.

The paper seeks to classify trading rationales small bits of text students ascribe to their paper trades when using an educational trading platform. The paper presents a study of "good" rationales ( thoughtful rationales ) vs poor rationales and based on expert assessments for them, several classifiers are developed, including CNNs and SVMs.

The paper is available as part of the ACL anthology and it is open access. Sadly, the data was confidential and the code is proprietary to Dr. Ying's employer.