Why Machines Cannot Learn Mathematics, Yet

Nowadays, Machine Learning (ML) is seen as the universal solution to improve the effectiveness of information retrieval (IR) methods. However, while mathematics is a precise and accurate science, it is usually expressed by less accurate and imprecise descriptions, contributing to the relative dearth of machine learning applications for IR in this domain. Generally, mathematical documents communicate their knowledge with an ambiguous, context-dependent, and non-formal language. Given recent advances in ML, it seems canonical to apply ML techniques to represent and retrieve mathematics semantically. In this work, we apply popular text embedding techniques to the arXiv collection of STEM documents and explore how these are unable to properly understand mathematics from that corpus. In addition, we also investigate the missing aspects that would allow mathematics to be learned by computers.

[1]  Abdou Youssef,et al.  Part-of-Math Tagging and Applications , 2017, CICM.

[2]  Petr Sojka,et al.  Software Framework for Topic Modelling with Large Corpora , 2010 .

[3]  William I. Grosky,et al.  Multi-sense embeddings through a word sense disambiguation process , 2019, Expert Syst. Appl..

[4]  Bruce R. Miller,et al.  Deep Learning for Math Knowledge Processing , 2018, CICM.

[5]  Nan Hua,et al.  Universal Sentence Encoder for English , 2018, EMNLP.

[6]  George A. Miller,et al.  WordNet: A Lexical Database for English , 1995, HLT.

[7]  Andrew Y. Ng,et al.  Improving Word Representations via Global Context and Multiple Word Prototypes , 2012, ACL.

[8]  André Greiner-Petter,et al.  Improving the Representation and Conversion of Mathematical Formulae by Considering their Textual Context , 2018, JCDL.

[9]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[10]  Ignacio Iacobacci,et al.  SensEmbed: Learning Sense Embeddings for Word and Relational Similarity , 2015, ACL.

[11]  Quoc V. Le,et al.  Distributed Representations of Sentences and Documents , 2014, ICML.

[12]  Ignacio Iacobacci,et al.  Embedding Words and Senses Together via Joint Knowledge-Enhanced Training , 2016, CoNLL.

[13]  S. Piantadosi.  Zipf’s word frequency law in natural language: A critical review and future directions , 2014, Psychonomic Bulletin & Review.

[14]  Nigel Collier,et al.  Towards a Seamless Integration of Word Senses into Downstream NLP Applications , 2017, ACL.

[15]  Simone Paolo Ponzetto,et al.  BabelNet: The automatic construction, evaluation and application of a wide-coverage multilingual semantic network , 2012, Artif. Intell..

[16]  Andrew McCallum,et al.  Efficient Non-parametric Estimation of Multiple Embeddings per Word in Vector Space , 2014, EMNLP.

[17]  Raymond J. Mooney,et al.  Multi-Prototype Vector-Space Models of Word Meaning , 2010, NAACL.

[18]  Nigel Collier,et al.  De-Conflated Semantic Representations , 2016, EMNLP.

[19]  Volker Markl,et al.  Semantification of Identifiers in Mathematics for Better Math Information Retrieval , 2016, SIGIR.

[20]  Frank Wm. Tompa,et al.  Multi-Stage Math Formula Search: Using Appearance-Based Similarity Metrics at Scale , 2016, SIGIR.

[21]  Daniel Jurafsky,et al.  Do Multi-Sense Embeddings Improve Natural Language Understanding? , 2015, EMNLP.

[22]  Michihiro Yasunaga,et al.  TopicEq: A Joint Topic and Mathematical Equation Model for Scientific Texts , 2019, AAAI.

[23]  Yue Yin,et al.  Preliminary Exploration of Formula Embedding for Mathematical Information Retrieval: can mathematical formulae be embedded like a natural language? , 2017, ArXiv.

[24]  Magdalena Wolska,et al.  Symbol Declarations in Mathematical Writing , 2010 .

[25]  Michael Kohlhase.  Math Object Identifiers - Towards Research Data in Mathematics , 2017, LWDA.

[26]  David M. Blei,et al.  Equation Embeddings , 2018, ArXiv.

[27]  Moritz Schubotz,et al.  Evaluating and Improving the Extraction of Mathematical Identifier Definitions , 2017, CLEF.

[28]  Jeffrey Pennington,et al.  GloVe: Global Vectors for Word Representation , 2014, EMNLP.

[29]  Ignacio Iacobacci,et al.  Embeddings for Word Sense Disambiguation: An Evaluation Study , 2016, ACL.

[30]  Luke S. Zettlemoyer,et al.  Deep Contextualized Word Representations , 2018, NAACL.

[31]  Tomas Mikolov,et al.  Enriching Word Vectors with Subword Information , 2016, TACL.

[32]  Hirokazu Anai,et al.  The Most Uncreative Examinee: A First Step toward Wide Coverage Natural Language Math Problem Solving , 2014, AAAI.

[33]  Roberto Navigli,et al.  A Unified Multilingual Semantic Representation of Concepts , 2015, ACL.

[34]  Giovanni Yoko Kristianto,et al.  Extracting Textual Descriptions of Mathematical Expressions in Scientific Papers , 2014, D Lib Mag..