Case Study:
Google’s understanding of the Health sector

May 2020

Executive Summary

Google uses Natural Language Processing (NLP) routines to convert its understanding of text into topics and concepts (Entities). We looked out how much understanding Google misses using this approach in the Health sector.

InLinks analyzed the page content of the top 10 Google results (in the UK Market) ranking for the phrase Kawasaki disease and compared the Named Entities recognized by Google’s NLP API with the proprietory routines designed by Inlinks to uncover gaps in Google’s machine learning in the Health sector

The results showed that 16.1% of entities seen on the results in the Health sector SERPS (search engine Results pages) were positively identified by Google.

This compares to 19.1% average across all analyzed industry sectors.

How the Sector Compares

Different sectors tend to be analyzed with a different degree of accuracy by the search engines. This stems from two main challenges.

  1. the more demand there is by consumers within a given sector, the more the need for search engines to apply more sophisticated entity recognition to better answer user queries.
  2. the more sophisticated the industry is in creating search-friendly content, the more search engines can surface topics and concepts.

 

Health Industry Google Analysis InLinks Analysis
Avg. Number of words per page 841
Avg. Number of Topics per page 7.4 46.1
Benchmark - Avg. nb of entities/page (all sectors) 9 48
Of which, Topic types are:    
- Persons 1 0
- Organizations 17 12
- Cities & geo. areas 16 14
- Concepts 18 245
Semantic Density   5.8

How the research was conducted

Google’s Search API returned URLs for the following sites competing for this phrase:

#bestpractice.bmj, #bbc, #chfed.org.uk, #gosh.nhs.uk, #nhs.uk, #societi.org.uk, #vasculitis.org.uk

The texts of each page are then sent to Google’s NLP API, in order to determine which entities are identified by the search engine. These are important for search since Google is then able to link these to its Knowledge Graph to feed its services including Google Discover, Google search, Voice Search and Google News. (Although, correct identification does not guarantee inclusion in these results)

Here is first of all the synthesis of the results returned by Google:

  • 18 concepts, including Kawasaki disease (detected 4 times) England (4) Coronavirus disease 2019 (2)
  • 14 geographical areas, including United Kingdom (5) Japan (3) Conwy (3)
  • 17 organizations, including National Health Service (2) Royal College of Paediatrics and Child Health (2) Agency for Science, Technology and Research (1)
  • 1 person, including Tomisaku Kawasaki (2)
  • 2 cities, including Shenzhen (1) Carlsbad, California (1)

Errors in Google’s Detection Rate

Here are some errors in the categorization of entities:

  • England categorized as a concept instead of geographical area
  • Nottinghamshire categorized as a concept instead of geographical area
  • Northern_Ireland categorized as a concept instead of geographical area
  • French_Alps categorized as a concept instead of geographical area
  • California categorized as a concept instead of geographical area

Most important entities (provided by InLinks), compared to those identified by Google:

  • Symptom (seen 9 times) => NOT detected by Google
  • Child (8) => NOT detected by Google
  • Complication (medicine) (7) => NOT detected by Google
  • Therapy (7) => NOT detected by Google
  • Technical support (6) => NOT detected by Google
  • Lip (6) => NOT detected by Google
  • Kawasaki disease (6) => detected by Google
  • Fever (6) => NOT detected by Google
  • Web page (5) => NOT detected by Google
  • United Kingdom (5) => detected by Google
  • Human eye (5) => NOT detected by Google
  • Medical diagnosis (5) => detected by Google
  • Disease (5) => NOT detected by Google
  • Medicine (5) => NOT detected by Google
  • Knee effusion (4) => NOT detected by Google

How can the Health Industry benefit from this report?

By understanding where Google is failing to recognize important concepts within the industry, there is an opportunity for companies in the sector to write clearer content that Google can better understand.

Another option is to explicitly state these concepts in Webpage schema for machine learning algorithms to take into account. This would require using Schema.org/WebPage and the "about" and mentions" properties for important concepts such as Symptom, Complication (medicine), Therapy, Lip, Fever, Human eye, Disease, Medicine, Knee effusion.

 

Internal linking of topics topages dedicated to each important topic will also help to reduce cannibalisation of content in Google’s understanding of contect within your content.

In Summary

Google's understanding of the Health market, based on its NLP algorithms remains limited at 16.1% for this industry. Businesses either need to improve their schema or make their content more understandable by Google to improve its level of understanding.

SERP results

https://www.nhs.uk/conditions/kawasaki-disease/

National Health Service, United Kingdom, Immunoglobulin therapy, England

https://www.gosh.nhs.uk/conditions-and-treatments/conditions-we-treat/kawasaki-disease

Great Ormond Street Hospital, Japan, United Kingdom

https://www.vasculitis.org.uk/about-vasculitis/kawasaki-disease

Kawasaki disease, Tomisaku Kawasaki, United Kingdom, Japan

https://www.nhs.uk/conditions/kawasaki-disease/causes/

National Health Service, South Korea, Japan, Asia

https://www.societi.org.uk/kawasaki-disease-and-covid-19/

Kawasaki disease, Coronavirus disease 2019, Royal College of Paediatrics and Child Health, Winthorpe, Nottinghamshire, Newark (parish), Nottinghamshire, England, Conwy

https://www.bbc.co.uk/news/health-52003804

Facebook, Twitter, Pinterest, WhatsApp, LinkedIn, Facebook Messenger, BBC News (TV channel), News, Scientific Advisory Group for Emergencies, England, China, University of Exeter, Conwy, Northern Ireland, United Kingdom, University College London, French Alps, Netherlands, Italy, Iceland, South Korea, Shenzhen

https://www.chfed.org.uk/how-we-help/information-service/heart-conditions/kawasaki-disease/

Tomisaku Kawasaki, Kawasaki disease, Cardiology, Fundraising Regulator, Warfarin, Nifedipine, Propranolol

https://www.societi.org.uk/

Kawasaki disease, United Kingdom, Coronavirus disease 2019, Strawberry, Rash, Ireland, Nottinghamshire, Newark (parish), Winthorpe, Nottinghamshire, England, Conwy, Coronavirus, Royal College of Paediatrics and Child Health

https://bestpractice.bmj.com/topics/en-gb/236

The BMJ, Medical diagnosis, C-reactive protein, United States, Carlsbad, California, California, BMJ (company), Agency for Science, Technology and Research, Government of Singapore

Find other market reports here

© 2019-2020 - InLinks.net - About us - Terms of Use - Privacy Policy