Searching for health information continues to be one of the most dominant areas of web search activity. Over the years, we've seen the development of better search capabilities in a newer generation of search engines that draw on semantic technologies (the meaning of language) to provide concept-based searching, in contrast to full-text search. Behind the scenes, the search engines draw on taxonomies and databases of medical concepts that include diseases, conditions, causes, symptoms, diagnoses, treatments, and other medically relevant attributes.
Healthline Networks has become one of the important technology providers for many of the health sites. David Bousfield, an analyst with Outsell, Inc., says that "Healthline Networks is quickly becoming a preferred supplier of semantic search, advertising, and content services for publishers, advertisers, and healthcare plan providers. It can also provide STM players with some ideas about the future of search."
Healthline's product portfolio is used by consumer media brands such as AARP, AOL, Health.com, iVillage's Your Total Health, and UnitedHealth Group's myOptumHealth to deliver comprehensive online health search results. Healthline also helps medical publishers such as ADAM, Gale Cengage, and StayWell simultaneously enrich their content, increase distribution, and improve monetization. For health plans such as Aetna, Healthline connects medical data and patient information to offer consumers richer, personalized health information while lowering the plan's cost of service.
Healthline built its semantic search engine in collaboration with doctors. According to Healthline, 75% of medical searches start with symptoms. So, the Healthline search site offers a SymptomSearch box, as well as TreatmentSearch (enter a health condition), DrugSearch (including a pill identifier by color, shape, and markings, and an interaction checker), and a DocSearch. Healthline's collection of tools, quizzes, and calculators can be used to help determine the risk for the most common conditions and diseases.
Yahoo! recently launched its new Yahoo! Health website, with a large library of health content, dynamic health information tools, and powerful search capabilities. It's powered by Healthline Networks. You'll notice the similarities-a symptom search, treatment search, drug search, etc. And, to really see the Healthline difference, run a search like "fever" in a general search engine like Google and then run it in Healthline's or Yahoo!'s symptom search. The guided navigation and filtering capabilities help users narrow results and hone in on specific information-no wading through endless pages of links.
HealthMash, developed by WebLib, combines sophisticated Web 2.0 search and discovery technology with semantic concepts in a simple yet highly informative user interface. The hybrid vertical search and metasearch is an innovative next generation semantic health search engine that is currently in public beta. The HealthMash developers have been working on projects with the National Library of Medicine (NLM) and the National Institutes of Health (NIH) for more than 5 years. HealthMash was first showcased at the Medical Library Association (MLA) annual meeting in 2009.
HealthMash uses a pragmatic mix of natural language processing tools, semantic engineering techniques, and multiple knowledge sources, including a proprietary Health Knowledge Base, to achieve both high precision and relevancy in its search results. The Health Knowledge Base is automatically generated from trusted health content sites and diverse knowledge sources, such as MeSH and UMLS (Unified Medical Language System). HealthMash also utilizes the web itself as a database. The Health Knowledge Base contains explicit knowledge about:
- Health Concerns
- Signs and symptoms
- Tests, procedures, treatments
- Drugs and substances and adverse effects
- Alternative, complementary and integrative medicine
Tamas Doszkocs, senior computer scientist at NLM, says the HealthMash search result page can be thought of as "a dynamic surrogate or distillation of the contents of many relevant heterogeneous information sources." In summary, he says, "HealthMash combines vertical semantic search of trusted health information, federated search (health news, videos, etc.), semantic clusters with mouse-over contexts for exploration and discovery (related concepts, health concerns, tests and treatments, etc.), and table of contents and topic clusters for drill down in search results and dynamic query modification."
The Health Knowledge Base is also available through an API or as a web service. The HealthMash search widget can be added to websites for free. WebLib can also build custom federated search solutions for medical libraries or other health organizations. The solution can include searching all of an organization's licensed and open access content. Customers for WebLib products include the NIH Library, WebFeat, Science.gov, and others.
Doszkocs notes that the current Health Knowledge Base-with some 60,000 concepts and millions of concept-pair associations-is the consumer health subset of a much more ambitious project. This will morph into a full Biomedical Knowledge Base with considerably more semantic content-more than 4 million concepts, with tens of millions of relationships in context.
Doszkocs says these are some other examples of consumer health search engines that employ semantic search capabilities: