Hello Everyone,
I did a literature survey to find the most interesting and useful papers which can help in improving the LB score. I hope this will be useful to the community.
List of Papers which I found most useful: – Many of these papers are from Amazon Search itself.
-
Graph-based Multilingual Language Model: Leveraging Product Relations for Search Relevance (by Amazon Science) – KDD’22 – Method developed to solve Task 2
-
ANTHEM: Attentive Hyperbolic Entity Model for Product Search – WSDM’22
-
Graph-based Multilingual Product Retrieval in E-Commerce Search
-
Improving Relevance Quality in Product Search using High-Precision Query-Product Semantic Similarity
-
Heterogeneous Network Embedding for Deep Semantic Relevance Match in
E-commerce Search
Summary:
-
Most of the papers use graph neural networks to get a better context of the product and the query.
-
Transformer models were adopted instead of the LSTM-based network.
-
One paper used Hyperbolic networks to model the power-law dynamics and scale-free nature of the product-query relations.
Other resources
- GitHub - NTMC-Community/MatchZoo: Facilitating the design, comparison and sharing of deep text matching models. – Repo containing the implementation of various Semantic Search algorithms in Tensorflow.
- GitHub - NTMC-Community/MatchZoo-py: Facilitating the design, comparison and sharing of deep text matching models. – Repo containing the implementation of various Semantic Search algorithms in PyTorch.
- Connected Papers can help search for similar papers.
- https://github.com/guyulongcs/Awesome-Deep-Learning-Papers-for-Search-Recommendation-Advertising – A repo containing papers related to product search.