Product Classification in E-Commerce using Distributional Semantics

Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 536–546, Osaka, Japan, December 11-17 2016. Product Classification in E-Commerce using Distributional Semantics Vivek Gupta , Harish Karnick Indian Institute of Technology , Kanpur {vgupta,hk}@cse.iitk.ac.in Ashendra Bansal , Pradhuman Jhala Flipkart Internet Pvt. Ltd., Bangalore {ashendra.bansal,pradhuman.jhala}@flipkart.com Abstract Product classification is the task of automatically predicting a taxonomy path for a product in a predefined taxonomy hierarchy given a textual product description or title. For efficient prod- uct classification we require a suitable representation for a document (the textual description of a product) feature vector and efficient and fast algorithms for prediction. To address the above challenges, we propose a new distributional semantics representation for document vector forma- tion. We also develop a new two-level ensemble approach utilizing (with respect to the taxonomy tree) path-wise, node-wise and depth-wise classifiers to reduce error in the final product classi- fication task. Our experiments show the effectiveness of the distributional representation and the ensemble approach on data sets from a leading e-commerce platform and achieve improved results on various evaluation metrics compared to earlier approaches. 1 Introduction Existing e-commerce platforms have evolved into large B2C and/or C2C marketplaces having large in- ventories with millions of products. Products in ecommerce are generally organized into a hierarchical taxonomy of multilevel hierarchical categories. Product classification is an important task in catalog formation and plays a vital role in customer oriented services like search and recommendation and seller oriented services like seller utilities on a seller platform. Product classification is a hierarchical classifi- cation problem and presents the following challenges: a) a large number of categories have data that is extremely sparse with a skewed long tailed distribution, b) a hierarchical taxonomy imposes constraints on activation of labels. If a child label is active then it is necessary for a parent label to be active, c) for practical use the prediction should happen in real time - ideally within few milli-seconds. Traditionally, documents have been represented as a weighted bag-of-words (BoW) or tf-idf feature vector, which contains weighted information about the presence or absence of words in a document by using a fixed length vector. Words that define the semantic content of a document are expected to be given higher weight. While tf-idf and BoW representations perform well for simple multi-class classi- fication tasks, they generally do not do as well for more complex tasks because the BoW representation ignores word ordering and polysemy, is extremely sparse and high dimensional and does not encode word meaning. Such disadvantages have motivated continuous, low-dimensional, non-sparse distribu- tional representations. A word is encoded as a vector in a low dimension vector space typically R100 to R300. The vector encodes local context and therefore is sensitive to local word order and captures word meaning to some extent. It relies on the ‘Distributional Hypothesis’(Harris, 1954) i.e. Similar words occur in similar contexts. Similarity between two words can be calculated via cosine distance between their vector representations. Le and Mikolov …

Content truncated to 500 words