We apply foundation models to data discovery and exploration tasks. Foundation models are large language models (LLMS) that show promising performance on a range of diverse tasks unrelated to their training. We show that these models are highly applicable to the data discovery and data exploration domain. When carefully used, they have superior capability on three representative tasks: table-class detection, column-type annotation and join-column prediction. On all three tasks, we show that a foundation-model-based approach outperforms the task-specific models and so the state of the art. Further, our approach often surpasses human-expert task performance. We investigate the fundamental characteristics of this approach including generalizability to several foundation models and the impact of non-determinism on the outputs. All in all, this suggests a future direction in which disparate data management tasks can be unified under foundation models.

Named entity recognition (NER) is a fundamental task for several important applications such as knowledge base construction and semantic search. So far, the focus has been on building machine learning models, which identify generic named entities (e.g., person, date). Such models can be used off-the-shelf without requiring ground truth labels for training. However, such models cannot generalize to specialized domains that have domain-specific named entities (e.g., the legal domain). In these cases, it is inevitable to generate ground truth data and experiment with a variety of models in order to achieve good performance. Motivated by a real use case from the financial sector, we discuss the approach and lessons learned when solving the NER problem in the legal domain. This task is particularly challenging because it requires extensive human expertise to produce high quality ground-truth labels. For solving the legal domain NER problem, we first crawl a large dataset of legal documents and then introduce a semi-automated process to generate high-quality labels for a set of eleven predefined named entities. We validate that the proposed approach achieves high quality labels that outperform popular out-of-the-box NER methods. On top of that, our method once followed, can generate ground truth labels for the pre-defined named entities for an unbounded number of documents. Next, we experiment with a set of models and training procedures and report their performance on the NER task. Our experimental evaluation confirms that most of the models can generalize very well, achieving F1-score between 86% and 98.9%. The dataset, the labels produced by human annotators and our semi-supervised approach, as well as our code are made available to the research community.

E-commerce applications rely heavily on session-based recommendation algorithms to improve the shopping experience of their customers. Recent progress in session-based recommendation algorithms shows great promise. However, translating that promise to real-world outcomes is a challenging task for several reasons, but mostly due to the large number and varying characteristics of the available models. In this paper, we discuss the approach and lessons learned from the process of identifying and deploying a successful session-based recommendation algorithm for a leading e-commerce application in the home-improvement domain. To this end, we initially evaluate fourteen session-based recommendation algorithms in an offline setting using eight different popular evaluation metrics on three datasets. The results indicate that offline evaluation does not provide enough insight to make an informed decision since there is no clear winning method on all metrics. Additionally, we observe that standard offline evaluation metrics fall short for this application. Specifically, they reward an algorithm only when it predicts the exact same item that the user clicked next or eventually purchased. In a practical scenario, however, there are near-identical products which, although they are assigned different identifiers, they should be considered as equally-good recommendations. To overcome these limitations, we perform an additional round of evaluation, where human experts provide both objective and subjective feedback for the recommendations of five algorithms that performed the best in the offline evaluation. We find that the experts’ opinion is oftentimes different from the offline evaluation results. Analysis of the feedback confirms that the performance of all models is significantly higher when we evaluate near-identical product recommendations as relevant. Finally, we run an A/B test with one of the models that performed the best in the human evaluation phase. The treatment model increased conversion rate by 15.6% and revenue per visit by 18.5% when compared with a leading third-party solution.

Recommender systems are an integral part of eCommerce services, helping to optimize revenue and user satisfaction. Bundle recommendation has recently gained attention by the research community since behavioral data supports that users often buy more than one product in a single transaction. In most cases, bundle recommendations are of the form "users who bought product A also bought products B, C, and D". Although such recommendations can be useful, there is no guarantee that products A, B, C, and D may actually be related to each other. In this paper, we address the problem of collection recommendation, i.e., recommending a collection of products that share a common theme and can potentially be purchased together in a single transaction. We extend on traditional approaches that use mostly transactional data by incorporating both domain knowledge from product suppliers in the form of hierarchies, as well as textual attributes from the products. Our approach starts by combining product hierarchies together with transactional data or domain knowledge to identify candidate sets of product collections. Then, it generates the product collection recommendations from these candidate sets by learning a deep similarity model that leverages textual attributes. Experimental evaluation on real data from the Home Depot online retailer shows that the proposed solution can recommend collections of products with increased accuracy when compared to expert-crafted collections.