Query Processing Language (QPL) Use Cases
The following case examples are based on recent customer implementations, and anonymized.
Technical Journal Publisher – Semantic Expansion
This publisher has been providing a subscription service for many of the top technical journals to academic and corporate medical and zoological research labs for more than a century. Part of their online strategy is to also offer single articles from their various journals for sale to the public.
After an analysis of search logs, it was discovered that many non-professionals using search on their main Web site were getting zero or a very low number of results. A key reason for this was because members of the public were not using the professional terms typically used in journal articles. For example, a query forSwine Flu returned just 3 articles, although there were more than 100 pertaining to the H1N1 virus. A search for seven year cicadas returned no results, yet 12 articles from the Journal of Insect Behavior discussed Cacadidae Magiciada, the scientific name for that creature.
QPL, the Query Processing Language was implemented to perform semantic expansion of terms and overcome this problem. QPL was programed to access a database containing biological scientific names mapped against a list of their common names. This existing resource was used to add synonyms and an OR operator, before the query was sent to the search engine. A non-professionals" search option was added to the UI.
Secondly, a “Did you Know?” feature was provided with search results, presenting a picture of the plant or animal together with the correct scientific name and other interesting facts about the species.
Overall sales of single articls increase by 10% following the addition of QPL to the system. The site is now popular with high school biology students, who will later in their careers become potential subscribers for the full service.
Pharmaceutical Development – Intellectual Property Control
You would be surprised how often the theft of IP is instigated by people inside the company. Typically this happens when employees have accepted another job or are leaving the company for other reasons.
This cutting edge pharmaceutical development company is exploiting new concepts in gene therapy. Competition is tough and any leak of IP related information is potentially harmful to the business.
In an effort to control this problem, the company has used their intranet search to help identify potential internal sources of IP loss. The intranet now uses QPL, a search engine independent Query Processing Language in the following way:
- When a user enters a search query, QPL analyzes it for specific terms related to sensitive projects
- QPL also knows who the searcher is (via their single-sign-on credentials)
- If the user does not have access permissions to see content relating to the search terms, then generally speaking, underlying document ACLs will prevent a security breach from happening via search.
But as an additional precaution, QPL sends an email to an IP Security Officer for potential further actions. If the user has access permission to see files that contain the search terms, a message may still be sent to the IP Security Officer if the terms used are particularly sensitive.
As a final line of defense, QPL imposes an additional security filter and over-rides the default user permissions, blocking access to the content.
After a quick review of the event, if it is determined that there is no potential for a security breach, the quarantined documents are emailed to the user with the message, “Your recent search using the terms X,Y,Z caused a temporary quarantine of the following documents, we are sorry for the delay, however the security of our information is critical to the success of our company.”
Other procedures are in place for when it is determined that the search was part of an actual attempt at IP theft.