The fifth edition of the Search and Data meetup is dedicated to e-commerce search and open source. A nice agenda to mark our return to the Meetup scene
The fifth edition of the Search and Data Meetup is dedicated to e-commerce search and open source. A nice agenda to mark our return to the Meetup scene.
For this upcoming meetup, we will have three talks and free discussions:
// René Kriegler, Director of E-commerce at OpenSource Connections, who will talk about implicit user feedback and how to use it to improve an e-commerce search engine. A presentation for data scientists but not only :-)
// Lucian Precup, CTO of all.site, who will present Elasticsearch and the adaptations to be made to integrate this open source technology into an e-commerce search engine.
// Jérémy Mésière, Manager of the Digital Factory team at Manutan, who will give a feedback on the integration of Elasticsearch in the future e-commerce search engines of Manutan International.
Find bellow some additional details about the agenda.
More than other domains, e-commerce search depends on implicit user feedback to optimise search result ranking. While search relevance is probably still a major criterion of search result quality in e-commerce, the user finally takes the buying decision based on criteria such as ‘an attractive price’ and ‘brand sympathy’ that are very hard to make explicit. On the other hand, this decision making can be observed implicitly in web tracking.
Unfortunately, e-commerce search cannot just use more generally known approaches to click modelling. Many of these models assume that the user would view the search results sequentially, top to bottom, while in e-commerce search we often use a grid layout which invites the user to go back and forth between results, especially given the heavy use of product imagery on the search result page. In addition, our model needs to consider contexts beyond the well-known position bias, such as the device type (mobile vs desktop) or the time of the day, together with the different types of signals (clicks, add-to-basket, checkout). Last but not least, many general click models ignore the fact that the number of observations can vary a lot between query-product pairs. Our greater uncertainty in the case of sparse events should be reflected in the model.
In this talk, I shall introduce an approach to using implicit user feedback that is based on Bayesian hierarchical modelling. It will provide a solution for dealing with position bias, including for grid layouts, and for dealing with further contexts, such as device types. The model will cope with varying quantities of observations and it allows to incorporate different types of events, such as clicks and checkouts.
René works as Director E-commerce at OpenSource Connections, empowering the company’s clients to make onsite search a key driver for improving their online sales.
He has worked in search for 15 years, including on projects for some of the top 10 German e-commerce sites. He is co-founder and co-organiser of MICES (Mix-Camp E-commerce Search), an event that brings together the e-commerce search community each year. His technological focus is on Solr, Elasticsearch and Lucene. He created and maintains the Querqy open source library for query rewriting. René is co-initiator of the Chorus project – an open source software stack that combines Querqy with other powerful tools to build e-commerce search and to measure and improve search quality.
Elasticsearch is an open source technology based on Apache Lucene which is the most widely used search engine library in the world. The Elasticsearch technology, and more generally the Elastic stack (ELK), is widely used and very well suited for use cases like Monitoring, Observability, Logs and Security. Areas such as Enterprise Search, Workplace Search, Site Search and especially E-Commerce Search require specific adaptations.
Lucian will present an extensive overview of about 30 points of attention when integrating Elasticsearch in an e-commerce context. Topics such as general architecture, auto-completion, relevance, facets, filters and other ways to refine searches, synonyms and taxonomies as well as personalization will be discussed.
A focus will be made on the e-commerce ecosystem and the modules, algorithms and utilities needed to have a complete, powerful and relevant software stack.
Lucian Precup, is the CTO of all.site, and develops solutions for indexing, searching and data analysis. Since 2010, Lucian built or helped building many e-commerce search engines among which Fnac.com and Carrefour.fr. His latest project, all.site, allows companies, universities and individuals organize information and make it accessible in the most relevant, secure and user friendly manner.
The Manutan Group is the European leader in BtoB e-commerce. Specialized in the distribution of equipment for businesses and local authorities, Manutan is constantly innovating to improve its customers’ experience. As part of its digital transformation, Manutan has launched several projects to overhaul its various systems. The most recent is a new search engine platform for the group’s e-commerce sites, including Manutan.fr. The new search engine is based on an open and modern software stack:
Jérémy Mésière is the Founder and Manager of the Digital Factory team at Manutan International. Created over a year ago, the Digital Factory team has set up a modern technical stack to serve the different business lines of the Manutan Group.
Dans cet article, nous expliquons comment nous sommes passés des anciens filtres de synonymes d'Elasticsearch aux nouveaux filtres de type graphe, les Synonym Graph Token Filter.
Dans cet article, nous présentons une méthode simple pour réécrire les requêtes utilisateurs afin qu'un moteur de recherche basé sur des mots clés puisse mieux les comprendre. Cette méthode est très utile dans le contexte d'une recherche vocale ou une conversation avec un chatbot, contexte dans lequel les requêtes utilisateur sont généralement plus verbeuses.
This article is a transcript of the lightning talk we presented this week at Haystack - the Search and Relevance Conference. We showed a method allowing to enrich and rewrite user queries using Wikidata and the Elasticsearch percolator.