NLP-Measuring semantic sentence similarity using baseline and neural models

The objective of this project is to improve the similarity between the sentences by implementing and designing various Machine Learning and Deep learning models in Natural Language Processing. By applying various techniques, we try to reduce the mean square error of the model and assess the distance between the words or sentences in the vector space using cosine distance similarity and word movers distance.

Language: Python, NLTK library

1. Pre-Requisites

• Python Editor (eg: Pycharm , Anaconda) with Python 3.

• Jupyter Notebook

• Pytorch (latest version) , GPU not necessary

• Working knowledge of Machine Learning, Deep Learning and NLP to start with

• Packages like NLTK, sklearn, Numpy, Pandas, Scipy, Sent2vec

• BioSentVec embedding for 4BioSentvec.ipynb is available in https://ftp.ncbi.nlm.nih.gov/pub/lu/Suppl/BioSentVec/BioSentVec_PubMed_MIMICIII-bigram_d700.bin

• Bio Word2Vec embedding for 3Word2Vec.ipynb is available in https://bio.nlplab.org

• The Dataset can be requested from https://portal.dbmi.hms.harvard.edu and https://sites.google.com/view/ohnlp2018/home(Task 2)

Research papers:

[1] Wang Y, Afzal N, Liu S, Rastegar-Mojarad M, Wang L, Shen F, Fu S, Liu H. Overview of the BioCreative/OHNLP Challenge 2018 Task 2: Clinical Semantic Textual Similarity. Proceedings of the BioCreative/OHNLP Challenge. 2018.

[2] Wang Y, Afzal N, Fu S, Wang L, Shen F, Rastegar-Mojarad M, Liu H. MedSTS: a resource for clinical semantic textual similarity. Language Resources and Evaluation. 2018, Jan 1:1-6.

[3] Chen Q, Peng Y, Lu Z. BioSentVec: creating sentence embeddings for biomedical texts. In: The 7th IEEE international conference on healthcare informatics; 2019.