Major Improvements Made to Searching in HeinOnline!

Customer Suggestion, Enhancements, Searching, What's New
Shannon Furtak

The Hein Company has long enjoyed a reputation of having the best customer service in the industry. We have always welcomed both positive and negative customer feedback. We recently surveyed our customers to find out what they liked best about HeinOnline and what features could use improvement. While we learned that we’re doing many things right, we found much that could be improved with our searching and search results.

We have taken this feedback seriously and will be implementing many of the suggestions we received as we work to enhance the HeinOnline interface over the next few months. In the meantime, we have drastically improved the algorithm we use for calculating search result relevancy.

Improvements to Search Relevancy Algorithm

We found that many users apply a “Google” search technique when using the advanced search boxes in HeinOnline. For example, when searching for a law review article, it is common to input a full or partial title, citation, and/or author name into the full text search box. Previously, the search result relevancy was based primarily on the frequency of term occurrence in document text; we have now boosted terms that occur in certain aspects of the metadata, such as the article title, citation, author, and publication title fields. We have also boosted term proximity: search terms that appear closer to one another will receive a higher relevancy rating.

Running the following search in the text search box in the Law Journal Library produces the following article as the first search result: whose eyes are you going to believe hoffman

Logically, the person entering this search was searching specifically for this article, and not for documents that contain all of the search terms. Search results will still include all documents that contain the terms entered, but because all of the words entered are found in the article title and author field, the correct result is now automatically ranked first in terms of relevancy. Previously, the exact same search produced the same number of results, but the desired article was result #25:

Like other large research databases and social media sites such as JSTOR, EBSCO, Twitter and LinkedIn, HeinOnline uses the robust Lucene/SOLR searching platform. The enhancements we have made allow for better search results for users who perform searches without using standard Lucene syntax. By analyzing hundreds of search strings and their corresponding results, our development team has been able to structure the search relevancy algorithm to predict the information users are most likely trying to obtain.

Additional Searching Enhancements

In addition to the improvements made to the relevancy algorithm, we have streamlined the appearance of available tools, such as PDF/Download and MyHein, and enhanced the search result display by automatically adding the first instance of matching text that occurs in each result:

Look for exciting information coming soon about HeinOnline’s new interface, which we expect to be ready this summer for AALL in Philadelphia! If you have feedback about these enhancements, or about HeinOnline generally, click the Feedback button in the upper right side of any screen in HeinOnline or contact us at (800) 277-6995,, or via LiveChat.