Natural Key Discovery in Wikipedia Tables

Bornemann, Leon; Bleifuß, Tobias; Kalashnikov, Dmitri V.; Naumann, Felix; Srivastava, Divesh

doi:10.1145/3366423.3380039

Cited by 12 publications

(3 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ODE layer in the generator transforms ⊕ , the concatenation of a noisy vector and a condition vector , into another latent vector ′ that will be fed into the generator (See Section 3.3). reasons, many web-oriented researchers focus on various tasks on tabular data [10,12,27,30,32,45,56,59,62,63]. In this work, generating realistic synthetic tabular data is of our utmost interest.…”

Section: ) (C)mentioning

confidence: 99%

See 1 more Smart Citation

OCT-GAN: Neural ODE-based Conditional Tabular GANs

Kim

Jeon

Lee

et al. 2021

Proceedings of the Web Conference 2021

View full text Add to dashboard Cite

Synthesizing tabular data is attracting much attention these days for various purposes. With sophisticate synthetic data, for instance, one can augment its training data. For the past couple of years, tabular data synthesis techniques have been greatly improved. Recent work made progress to address many problems in synthesizing tabular data, such as the imbalanced distribution and multimodality problems. However, the data utility of state-of-the-art methods is not satisfactory yet. In this work, we significantly improve the utility by designing our generator and discriminator based on neural ordinary differential equations (NODEs). After showing that NODEs have theoretically preferred characteristics for generating tabular data, we introduce our designs. The NODE-based discriminator performs a hidden vector evolution trajectory-based classification rather than classifying with a hidden vector at the last layer only. Our generator also adopts an ODE layer at the very beginning of its architecture to transform its initial input vector (i.e., the concatenation of a noisy vector and a condition vector in our case) onto another latent vector space suitable for the generation process. We conduct experiments with 13 datasets, including but not limited to insurance fraud detection, online news article prediction, and so on, and our presented method outperforms other state-of-the-art tabular data synthesis methods in many cases of our classification, regression, and clustering experiments. CCS CONCEPTS• Computing methodologies → Machine learning; Neural networks.

show abstract

Section: ) (C)mentioning

confidence: 99%

“…reasons, many web-oriented researchers focus on various tasks on tabular data [10,12,27,30,32,45,56,59,62,63]. In this work, generating realistic synthetic tabular data is of our utmost interest.…”

Section: ) (C)mentioning

confidence: 99%

OCT-GAN: Neural ODE-based Conditional Tabular GANs

Kim

Jeon

Lee

et al. 2021

Proceedings of the Web Conference 2021

View full text Add to dashboard Cite

show abstract

“…For example, work on identifying data transformations can help with increasing the joinability of tables [21,27,50]. Improvements to discovery engines [4,9,36,54], and key detection approaches [7,11] help with inferring higher quality schemas.…”

Section: Related Workmentioning

confidence: 99%

Niffler: A Reference Architecture and System Implementation for View Discovery over Pathless Table Collections by Example

Gong¹,

Zhu²,

Galhotra³

et al. 2021

Preprint

View full text Add to dashboard Cite

Identifying a project-join view (PJ-view) over collections of tables is the first step of many data management projects, e.g., assembling a dataset to feed into a business intelligence tool, creating a training dataset to fit a machine learning model, and more. When the table collections are large and lack join information-such as when combining databases, or on data lakes-query by example (QBE) systems can help identify relevant data, but they are designed under the assumption that join information is available in the schema, and do not perform well on pathless table collections that do not have join path information.We present a reference architecture that explicitly divides the end-to-end problem of discovering PJ-views over pathless table collections into a human and a technical problem. We then present Niffler, a system built to address the technical problem. We introduce algorithms for the main components of Niffler, including a signal generation component that helps reduce the size of the candidate views that may be large due to errors and ambiguity in both the data and input queries. We evaluate Niffler on real datasets to demonstrate the effectiveness of the new engine in discovering PJ-views over pathless table collections.

show abstract