Rapport de recherche sur l'état de la recherche sur site dans l'enseignement supérieur avec The Chronicle of Higher Education | Télécharger le rapport
Rapport de recherche sur l'état de la recherche sur site dans l'enseignement supérieur avec The Chronicle of Higher Education | Télécharger le rapport

Jan. 31, 2022

Tom Humbarger

|

3 min de lecture

SearchStax has released new AI-powered and machine learning features for SearchStax Site Search. The AI-powered site search capabilities leverage machine learning to increase engagement when a user is conducting searches on a website. Artificial intelligence currently powers the Auto-Suggest and Related Searches features.

AI-Powered Auto-Suggest Searches

Suggestion automatique enables real-time suggestions when the user starts typing a search term. SearchStax Site Search uses machine learning to leverage search history to build recommendations of searches that users may be looking for. This improves findability, reduces spelling errors and helps direct users to the right content. 

There is also an Add Word feature that lets administrators pre-seed the Auto-Suggest suggestions with their own tips until the auto-suggest feature builds a sufficient database of suggestions from historical searches. On the Australian Catholic University website, powered by SearchStax Site Search, the auto-suggest feature kicks in after typing a single letter.

Auto-Suggest feature in action on the Australian Catholic University website

AI-Powered Related Searches

Recherches associées enables Google-like search suggestions that users expect to see at the bottom of their search results. This lets users further refine their searches and enhances discovery. Using learnings driven by user interactions and correlation of searches, the AI-driven capability makes it easy for business users to offer a better experience for users. AI driven by historical searches, SearchStax Site Search will offer up to 10 related search terms that may be of further interest to the user. Clicking on a related search term will launch a new search for that search phrase.

Administrators can also manually add Related Search results individually or uploaded by a file through the SearchStax Site Search interface. When Related Searches have been manually added, they appear at the front of the Related Search results. The AI-Powered search results are then added after the manually added results. In the example below, we have identified 5 related searches that will appear in the search results whenever a user enters ‘sitecore’ on our search page.

Related Searches feature in SearchStax Site Search

When a user enters ‘sitecore’ as a search term in a search on our website, the user is presented with 10 related search terms at the bottom of the search results. The first 5 results were added manually using the SearchStax Site Search interface as displayed above and the last 5 results are powered by artificial intelligence. The combination of manual and AI-driven results will give marketers the ability to deliver the most relevant and personalized results for their customers.

Example of AI-Driven and manually curated Related Searches – the shaded results are driven by AI

AI-Powered Related Searches

SearchStax has also released a new SearchStax Site Search API that retrieves Related Searches defined for a specific search phrase in SearchStax Site Search. The API can be accessed through any tool that assembles HTTP requests and the result is a Json document of related-search strings for display on a search page. Existing SearchStax Site Search users will be able to quickly add AI-powered Related Search to their websites by using the new API and the Application d'interface utilisateur de recherche.

Recherche sur le site SearchStax gives digital marketers insights into search and control over the search experience. SearchStax Site Search gives marketers the power to leverage search analytics while easily creating a more modern search experience with easy-to-use tools. Contact us for a demo and evaluation to see how we make powerful search easy.

Par Tom Humbarger

Responsable principal des programmes de marketing

« Le temps de vos développeurs est précieux. Ils devraient se concentrer sur des activités à valeur ajoutée, comme l'amélioration des expériences de recherche, plutôt que sur les détails opérationnels fastidieux du déploiement, de la gestion et de la mise à l'échelle de l'infrastructure Solr… »

Vous aimerez peut-être aussi :

fr_CAFrançais du Canada