Open AccessAuto-generation of the customer questions and their ranking in e-commerce system
Author(s) -
Yulia Dyulicheva
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1399/3/033081
Subject(s) - ranking (information retrieval) , bayesian probability , computer science , key (lock) , e commerce , task (project management) , machine learning , artificial intelligence , stage (stratigraphy) , data mining , information retrieval , world wide web , engineering , paleontology , computer security , systems engineering , biology
The paper proposes the framework for auto-generation of the customer questions to help customers make choice of the best products based on their needs. The task of the optimal search organisation from millions of products is crucial for e-commerce systems. We propose the approach based on the following stages: the web-scraping stage of products reviews sites, pre-processing stage to detect key-phrases based on TextRank algorithm and POS-tagging, the subjective probabilities estimation stage to detect questions estimations and their ranking based on TextRank algorithm and Bayesian rule.