E-Commerce Product Categorization via Machine Translation

Abstract

E-commerce platforms categorize their products into a multi-level taxonomy tree with thousands of leaf categories. Conventional methods for product categorization are typically based on machine learning classification algorithms. These algorithms take product information as input (e.g., titles and descriptions) to classify a product into a leaf category. In this article, we propose a new paradigm based on machine translation. In our approach, we translate a product's natural language description into a sequence of tokens representing a root-to-leaf path in a product taxonomy. In our experiments on two large real-world datasets, we show that our approach achieves better predictive accuracy than a state-of-the-art classification system for product categorization. In addition, we demonstrate that our machine translation models can propose meaningful new paths between previously unconnected nodes in a taxonomy tree, thereby transforming the taxonomy into a directed acyclic graph. We discuss how the resultant taxonomy directed acyclic graph promotes user-friendly navigation, and how it is more adaptable to new products.

Explainable Automatic Industrial Carbon Footprint Estimation From Bank Transaction Classification Using Natural Language Processing

“…304 verbs such as "to be" and "do") is based on the Spanish stop 327 word list from the NLTK Python library 13 . For tokenizing 328 purposes, the same NLTK Python library 13 was used and the 329 resulting tokens were lemmatized with the spaCy Python 330 library 14 using the es_core_news_sm model 15 . For the LSTM model, the Tokenizer 18 function from the Keras Python library was used.…”

“…Many previous studies have applied ML in fields such as Ecommerce [15], incident management in information systems [16], and medical record analysis [17]. In finance [18], [19], ML models have been considered for detecting financial opportunities in social networks [20], fraud [21], [22], market sentiment [23], risk [24], accounting [25], and financial transaction classification [26].…”

“…Stop word removal (including general-usage verbs such as "to be" and "do") is based on the Spanish stop word list from the NLTK Python library 13 . For tokenizing purposes, the same NLTK Python library 13 was used and the resulting tokens were lemmatized with the spaCy Python library 14 using the es_core_news_sm model 15 .…”

“…Available at https://www.nltk.org, November 2022 14. Available at https://spacy.io, November 2022 15. Available at https://spacy.io/models/es#es_core_news_sm, November 2022 16.…”

Automatic Categorization of Multi Marketplace FMCGs Products using TF-IDF and PCA Features

“…Several studies related to categorizing a product have been carried out with the use of machine learning. For example, research on the categorization of e-commerce products is carried out by proposing a machine translation paradigm using classification techniques by looking at nodes in the taxonomic tree [6]. Meanwhile, Chavaltada et al compared the performance of various machine learning techniques on product categorization in the proposed framework [7].…”