(NeurIPS D&B 2024) STaRK: Benchmarking LLM Retrieval on Textual and Relational Knowledge Bases
-
Updated
Feb 6, 2026 - Python
(NeurIPS D&B 2024) STaRK: Benchmarking LLM Retrieval on Textual and Relational Knowledge Bases
UrnaDB is a NoSQL database support diverse data types and transactions.
Refinery is a tool to extract and transform semi-structured data from Excel spreadsheets of different layouts in a declarative way.
Pattern matching language for transforming complex tree-structured data (strings, XML, JSON, algebraic expressions). Built-in symbolic algebra, backtracking, and exact arithmetic. Portable C99.
A dataset for extracting information from repair manuals
Implementation of the semi-structured inference model in our ACL 2020 paper, INFOTABS: Inference on Tables as Semi-structured Data.
Repository containing code for the NAACL 2021 paper (Incorporating External Knowledge to Enhance Tabular Reasoning)
An ActiveModel extension to model your semi-structured data using embedded associations
Urban Dict spelling variant dataset. Source code of How to Evaluate Word Representations of Informal Domain?
This repository contains the official code for the paper : Realistic Data Augmentation Framework for Enhancing Tabular Reasoning (Findings-EMNLP, 2022).
Coherent data analysis library
A semi-automatic web-based annotation tool for MyFixit dataset :
Schema inference for semistructured data using Formal Concept Analysis
Framework to manipulate semi structured documents and extract data from them
Implementation of the semi-structured inference model in our ACL 2023 paper: INFOSYNC: Information Synchronization across Multilingual Semi-structured Tables.
Java Standalone application for querying XML documents with requests with preferences (GTPs requests with preferences)
Web-based workflow management system that computes candidate tool workflows given input file(s) and the user's requirements regarding the output. Afterwards, runs a workflow selected by the user from the list of candidates. Implemented in Bracmat (~75%) and Java (~25%).
An open collection includes 100+ semi-structured textual datasets. (LOG datasets, TXT datasets, CSV datasets etc.)
Extracts key release details from unstructured text to create clear, structured summaries.
Python binding of Archery
To associate your repository with the semi-structured-data topic, visit your repo's landing page and select "manage topics."