Custom web search


Search engine is the most important and must-to-know thing nowadays. It lets you to search over the internet to get the specific search results. At the same time, it's a skill to ask right query to get the right results out of it. Many times it happens that you are searching for something on a particular search engine but end up getting totally different results. This may not be search engine’s fault but the query we asked for is not sufficient enough to get the desired results. For eg: When we are searching for the Paris tourism data or places with the query keyword -  Paris, it might show you the result for  Paris Hilton. On another note, if we want to have very specific results coming out of search engine, we need to refine queries accordingly and make it very specific.

How It Works:

We were working on a Website project which had a requirement to get the traveling related data only. We had given a text box to allow end user to search. But again, if it is a plain text search without filter, it will display irrelevant results just like above example. So we have used different APIs to make the relevant search as follows:

Videos - YouTube, dailymotion
Images - Flicker, photobucket, picasa
Web - Google API
Blog - tripadvisor, onelyplanet

Along with these search engine APIs, we added some predefined filters which will make the search more relevant. Also, we added keyword like trip, travel, places to be appended along with the entered query. This won’t let user know that their query is getting refined for getting detailed results. It will look like user is adding any words but getting the desired results. The API filter the data and give only data related to travel. We added the categorization of the content for good user experience here we broadly define like search for only Images, Videos, Blogs.


This custom web search was very important for the website as it was a good user experience to an end user. Below are the benefits we got out of this implementation:
  • Improve relevancy of data.
  • Easily available data pulled from different sources.
  • Data categorization.
  • No additional effort needed from end users.
  • Good user experience.


Michael Patterson sat down with the CEO of Boston Byte, Mustapha Shaikh to discuss the significance and rapid digitization of the healthcar...