For the last few months, our engineers have been busy behind the scenes re-structuring our search capability to deliver what we believe is an excellent scholarly publishing search experience. Our new search engine is significantly faster than before and returns more relevant results with many ways to narrow, broaden, filter or sort. You can see it in action on this Malaria search.
For years, scholarly publishers have lagged behind in providing similar search capabilities to, for example, Borders or Expedia, but not any more! Now PLoS is helping the scientific community – the birthplace of the internet – to make up some lost ground by offering what we believe is the most powerful and flexible search functionality available from any scientific and medical publisher.
Those of you who like technology, will be interested to know that we have also re-organized our infrastructure for better search performance. We are now using Solr, an Open Source search server originally developed by CNET Networks, with the Lucene Open Source search engine. A separate server is dedicated to our search engine to efficiently fetch information from our library of articles without impacting journal performance.
So what's the benefit of all this hard work? Users now have more search choices, better results and faster service whether they choose simple or advanced search. To take search to the next level we had to think about what options are important to our users when they search, and then provide this information up front, which means less effort for our users.
Now, when you do a ‘simple’ or ‘advanced’ search on any PLoS journal, your search will return a list of all PLoS articles from that journal, but you will additionally see a whole range of ways to narrow; broaden; filter or sort your results. Here's just a flavor or what you can do with our new search engine:
• View some key Article-Level Metrics of the search results
• Narrow search results by subject categories related to your results
• Sort by relevance, date and most popular based on article usage statistics
• Discover what article types are included in your search (research article, essay etc)
• Find out in which section of the article your keywords appear (abstract, discussion etc)
• Refine your search as you go, adding or removing different parameters from your search query
• Quickly make related searches such as related Authors, Academic Editors, and Institutions
• Find a specific article by DOI , the Year/Volume or eNumber
• Use the large free text box to construct complicated searches
It's important to say that the development of search is an iterative process and future releases will provide even more options such as searching on more elements of our data (for example, a wider range of our Article-Level Metrics data).
Here's what Sara Wood, Director of Product Strategy for PLoS has to say about these improvements:
"When an organization reaches a critical mass of content such as PLoS (or at my previous organization Flickr, the photo sharing site owned by Yahoo!), it's vitally important that users can reach the content they need consistently. When I joined PLoS, I made it a priority to improve search to help users find the content they need when they need it and I hope that today's release does just that. I would welcome user feedback to email@example.com."
Here's what the rest of us have to say about these improvements: “wow, enjoy”!