Semantic search

Analyzing the problem

  • Visitors can't find the product they are looking for
  • Regular keyword searches don't bring the correct results in case of a typo / mistake
  • Websites don't have a good order by/ filter option on the result page
  • Regular search algorithms can't deal with synonyms
  • Websites don't provide relevant alternatives when confronted with 0 results

The importance of a search

  • MarketingSherpa data reveals that almost half of the visitors who land on a website go immediately to the search function. Customers who use the search box on ecommerce sites convert at nearly three times the rate of general browsers.
  • According to WebSideStory. "More than 80% of all Internet-based purchases involve a search for products within a website, making an effective site search solution a key success factor in online retailing"
  • In an eyetracking study, the usability guru Jakob Nielsen described 57% of the users of a web shop as search dominant, meaning that they were by nature drawn to and highly likely to use the search tool of the website.

Goal

  • Find more relevant results for any given text search request entered by a user. This feature attempts to interpret the user's request (within the boundaries of the travel domain) in order to provide more relevant product results than a standard key-word search.

Challenge

  • Searching through all available content (structured and unstructured) aggregated in a vertical search structure (entity levels).
  • Better interpretation of search queries (including functionality for spelling corrections, synonym replacements) and understanding the context (semantics) of the (natural language) search query.
  • Setting up a model in which it is possible to define the meaning of (and relationship between) terms used in the travel domain.
  • Facilitating optimal user guidance (guided / facetted navigation and sorting) and delivering great user experience (repeating visitors).

Solution

TourWeb contains a collection of components and tools that can enhance the search capabilities of a website built on top of an Endeca-based index. TourWeb can "understand" a variety of travel related questions asked by website users (through the implementation of semantic search techniques) which produces better and more relevant search results when compared to the normal keyword search.

  • Flexible query interpretation
  • Implement stemming, synonym replacements, allow custom definition of words, 'fuzzy' key-word matching
  • Create an ontology (limited to the travel domain)

What's next?

Glossary

"Semantic search seeks to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable dataspace, whether on the Web or within a closed system, to generate more relevant results." [ Wikipedia ]
"A vertical search engine, as distinct from a general web search engine, focuses on a specific segment of online content. The vertical content area may be based on topicality, media type, or genre of content." [ Wikipedia ]
TourWeb E-Business Suite External Data / Systems Data / Systems Integration Partner Services Various Interfaces