-
ECIR 20242024We investigate the integration of Large Language Models (LLMs) into query encoders to improve dense retrieval without increasing latency and cost, by circumventing the dependency on LLMs at inference time. SoftQE incorporates knowledge from LLMs by mapping embeddings of input queries to those of the LLM-expanded queries. While improvements over various strong baselines on in-domain MS-MARCO metrics are
-
ACM CHIIR 20242024Product returns are an increasing environmental problem, as an estimated 25% of returned products end up as landfill [10]. Returns are expensive for retailers as well, and it is estimated that 15-40% of all online purchases are returned [34]. The problem could be mitigated by identifying issues with a product that are likely to lead to its return, before many have sold. Understanding and predicting return
-
CHIIR 20242024To help customers who are still in the exploration phase, Web search engines and e-commerce websites often provide relevant Q&As in widgets, such as ‘People Also Ask’ and ‘Customers Also Ask Alexa’, with additional information. In this work, we propose to enrich this customer experience by rendering related products under each Q&A based on an automated online query recommendation. We define what are the
-
2024The traditional cascading Entity Resolution (ER) pipeline suffers from propagated errors from upstream tasks. We address this issue by formulating a new end-to-end (E2E) ER problem, Signal-to-Entity (S2E), resolving query entity mentions to actionable entities in textual catalogs directly from audio queries instead of audio transcriptions in raw or parsed format. Additionally, we extend the E2E Spoken Language
-
ECIR 20242024Off-policy evaluation (OPE) methods allow us to compute the expected reward of a policy by using the logged data collected by a different policy. However, when the number of actions is large, or certain actions are under-explored by the logging policy, existing estimators based on inverse-propensity scoring (IPS) can have a high or even infinite variance. Saito and Joachims [13] propose marginalized IPS
Related content
-
February 28, 2022Novel pretraining method enables increases of 5% to 14% on five different evaluation metrics.
-
February 18, 2022Method using hyperboloid embeddings improves on methods that use vector embeddings by up to 33%.
-
January 19, 2022The scientist's work is driving practical outcomes within an exploding machine learning research field.
-
January 14, 2022A new metric-learning loss function groups together superclasses and learns commonalities within them.
-
December 14, 2021Two NeurIPS papers examine the assignment of the same label to multiple categories, fast training of Transformer-based models.
-
November 19, 2021Identifying descriptions of events that did not take place in product reviews improves product retrieval results.