Search and information retrieval
Microsoft Research Blog
RASCAL: Novel robotics for scalable and highly available automated storage and retrieval
RASCAL is an untethered robot with a modular design, allowing it to move flexibly along and between evenly spaced storage shelves. Discover how it can address the availability and scalability challenges of existing automated storage…
Publication
Towards Group-aware Search Success
Dataset Source Code
ORCAS: Open Resource for Click Analysis in Search
ORCAS is a click-based dataset associated with the TREC Deep Learning Track. It covers 1.4 million of the TREC DL documents, providing 18 million connections to 10 million distinct queries.
Dataset Source Code
TREC Tip-of-the-Tongue Track
Tip-of-the-tongue (ToT) known-item retrieval is defined as “an item identification task in which the searcher has previously experienced an item but cannot recall a reliable identifier” (i.e., “It’s on the tip of my tongue…”). The…
Dataset Source Code
TREC Deep Learning Track
The TREC Deep Learning Track studies information retrieval in a large training data regime. This is the case where the number of training queries with at least one positive label is at least in the…
GitHub Publication Publication Publication Publication Publication Publication Publication Publication Publication
Project
Project GraphRAG
LLM-Derived Knowledge Graphs GraphRAG (Graphs + Retrieval Augmented Generation) is a technique for richly understanding text datasets by combining text extraction, network analysis, and LLM prompting and summarization into a single end-to-end system. GraphRAG and…
Microsoft Research Blog
Research Focus: Week of April 1, 2024
In this issue: New research helps COMET embrace African languages; FeatUp improves deep features, a computer vision research cornerstone; LLMs in the Imaginarium: Tool Learning through Simulated Trial and Error; Benchmarking LLMs across languages and…
Publication