We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.IR

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Learning Representations for Zero-Shot Retrieval over Structured Data

Authors: Harsh Kohli
Abstract: Large Scale Question-Answering systems today are widely used in downstream applications such as chatbots and conversational dialogue agents. Typically, such systems consist of an Answer Passage retrieval layer coupled with Machine Comprehension models trained on natural language query-passage pairs. Recent studies have explored Question Answering over structured data sources such as web-tables and relational databases. However, architectures such as Seq2SQL assume the correct table a priori which is input to the model along with the free text question. Our proposed method, analogues to a passage retrieval model in traditional Question-Answering systems, describes an architecture to discern the correct table pertaining to a given query from amongst a large pool of candidate tables.
Comments: 11 pages, 3 figures, 2 tables
Subjects: Information Retrieval (cs.IR)
Cite as: arXiv:2111.00123 [cs.IR]
  (or arXiv:2111.00123v1 [cs.IR] for this version)

Submission history

From: Harsh Kohli Mr. [view email]
[v1] Fri, 29 Oct 2021 23:43:19 GMT (347kb,D)

Link back to: arXiv, form interface, contact.