Search

Main blog

Welcome to our Main Blog, the cornerstone of our exploration into information retrieval. This dedicated space serves as a comprehensive repository where we delve into our research, findings, and various topics predominantly centered around information retrieval.

Apache Solr Learning To Rank Feature Extraction and qTime

Apache Solr Learning To Rank Feature Extraction and qTime

How a learning to rank query works in Solr? How we can obtain the required features extraction time from the Solr qTime parameter?
Using BERT To Improve Search Relevance

From Training to Ranking: Using BERT to Improve Search Relevance

Learning about text ranking using Deep Learning with BERT transformer. From training to neural re-ranking, with code snippets and examples.
Artificial Intelligence Applied to Search Introduction

Artificial Intelligence Applied to Search: Introduction

How does Artificial Intelligence impact Search? This post explores the state of the art of AI applied to Information Retrieval in Open Source.
A Learning to Rank Project on a Daily Song Ranking Problem - Sease

A Learning to Rank Project on a Daily Song Ranking Problem – Part 4

Query-level features and under-sampled queries, how to handle them? Find it out, with our new Learning to Rank implementations
Apache Solr sow Parameter (Split on Whitespace) and Multi-Field Full-Text Search

Apache Solr sow Parameter (Split on Whitespace) and Multi-Field Full-Text Search

This blog post explores the Apache Solr multi-field search limitations with a focus on the sow(split on whitespace) parameter.
ltr project

A Learning to Rank Project on a Daily Song Ranking Problem – Part 3

This blog post aims to illustrate how to generate the query Id and how to manage the creation of the Training Set
Several analyses on a Learning to Rank model and explanation of its behavior using the open-source library SHAP.

A Learning to Rank Project on a Daily Song Ranking Problem – Part 2

This blog post is about several analysis on a LTR model and its explanation using the open source library SHAP
Elasticsearch _source, doc_values and store Performance

Elasticsearch _source, doc_values and store Performance

In this blog post, the elasticsearch _source field is compared with stored fields and docvalues from a performance point of view
Offline Search Quality Evaluation Rated Ranking Evaluator (RRE)

Offline Search Quality Evaluation: Rated Ranking Evaluator (RRE)

The Rated Ranking Evaluator (RRE) is an offline search quality evaluation library for both Apache Solr and Elasticsearch.
learning to rank project

A Learning to Rank Project on a Daily Song Ranking Problem

This blog post aims to illustrate step by step a Learning to Rank project on a Daily Song Ranking problem using open source libraries.
Apache Solr Learning To Rank Interleaving

Apache Solr Learning To Rank Interleaving

Interleaving is an online evaluation approach for ranking functions, contributed to Apache Solr Learning to Rank by Sease.
If you're struggling with music information retrieval, don't worry - we're here to help! Our team offers expert services and training to help you optimize your search engine and get the most out of your system. Contact us today to learn more!

Music Information Retrieval: the Intervals Matrix

In this post we describe what is an Intervals Table and how to build it using a Behaviour-Driven-Development (BDD) approach.
Explaining Learning to Rank Models with Tree Shap

Explaining Learning to Rank Models with Tree Shap

Explainability and Interpretability of Learning To Rank models are vital in Information Retrieval, in this blog we present Tree SHAP.
Secrets of Interleaving approaches for Learning To Rank online testingevaluation. It includes implementation details and procons analysis

Online Testing for Learning To Rank: Interleaving

Secrets of Interleaving approaches for Learning To Rank online testing/evaluation. It includes implementation details and pro/cons analysis.
The Importance of Online Testing in Learning to Rank – Part 1

The Importance of Online Testing in Learning to Rank – Part 1

It is fundamental to online test your Learning To Rank system, this blog shows you how it can be implemented and the most common mistakes.
DocValues VS Stored Fields : Apache Solr Features and Performance SmackDown

DocValues VS Stored Fields: Apache Solr Features and Performance SmackDown

This blog post aims to explain Docvalues and Store fields in Apache Solr for operations in which they can be used interchangeably.
weighted synonyms

Introducing Weighted Synonyms in Apache Lucene/Solr

This blog post is about our latest contribution to the Apache Lucene project: introducing weighted synonyms to provide better query expansion.
music information retrieval

Music Information Retrieval: the Intervals Table

In this post we describe what is an Intervals Table and how to build it using a Behaviour-Driven-Development (BDD) approach.