Unsupervised Learning for Product Ontology from Textual Reviews on E-Commerce Sites
| Autor: | Xuan Sun, Longquan Jiang, Minghuan Zhang, Cheng Wang, Chen Ying |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Computer science
business.industry Sentiment analysis Bootstrapping (linguistics) Ontology (information science) Product type computer.software_genre Categorization Feature (machine learning) Unsupervised learning Product (category theory) Artificial intelligence business computer Natural language processing |
| Zdroj: | Proceedings of the 2019 2nd International Conference on Algorithms, Computing and Artificial Intelligence. |
| Popis: | On modern e-commerce sites, textual reviews are rich in sentiment and opinions about a product, in particular, its specific attributes. Product ontologies, consisting of a taxonomic categorization of lists of such attributes and product types, are useful for a wide range of applications, such as opinion mining and sentiment analysis. Unfortunately, with a paucity of fine-grained hierarchical categorization of products, many smaller sites feature only coarse high-level categories. We present the Iterative Bootstrapping Process (IBP), an unsupervised learning method for such fine-grained hierarchical categorization of products using coarse categories from Chinese product textual reviews. Results show that our model can extract useful product attributes and still can achieve a high accuracy on the task for categorizing unseen products. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|