Products :: Related Searches
Sematext's Related Searches products offers the familiar
"People Who Searched For This, Also Searched For..."
functionality. When a person runs a query X, the Related
Searches service offers several other queries other people who
run the same X query have performed. You might have seen this
functionality on Google's search results pages, and possibly in
commercial enterprise search vendors' offerings.
Business Value / Benefits
- Better search experience
- Increased sales for ecommerce sites
- Increased ad sales due to increased page views per visit, increased visit frequency, and increased time on site
Do You Need It?
How do you determine if Sematext's Related Searches is for you?
- Your service/business benefits from people discovering and finding content or products you offer
- Your service/business loses revenue if people don't quickly find what they are looking for
Integration with any web application is seamless and consists of
- Search Beacon
- Related Searches Service
The Search Beacon is a simple servlet that collects query
information such as query strings, session IDs, user IDs, etc.
It is integrated into pages as a pixel-sized image completely
invisible to users.
The Related Searches Service is a simple servlet with a REST
API. Given the query string, the service returns a list of
related queries, ordered by the strength of their relatedness to
the input query. To display related searches in the search
results pages, the web appplication that executes user queries
typically communicates with this service, gets related searches,
and passes them to the user interface layer for display.
Q: Does Related Searches product work
only with Lucene or Solr ?
A: No, Sematext's Related Searches is
vendor neutral - it works with Lucene, Solr, Nutch, Google Search
Appliance, Endeca, FAST ESP, Autonomy, Attivio, Vivismo,
Q: When are related searches computed? In
A: Related searches are (re)computed
periodically - in batch. The frequency of recomputation can
range anywhere from every few minutes to weekly, depending on
the nature, need, and scale of the application.
Q: How is search relatedness computed ?
A: Please get in touch if you would like to
find out more about our secret sauce.