Contextual Advertising & Semantic Targeting

• Contextual advertising uses location, age, birthday, education, and viral connections as a way to target consumers
• Semantic targeting relies on concepts and relationships rather than keywords or phrases
• Semantic advertising systems rely on precise automation, real-time analysis, and monitoring socio-cultural connections
• Social Networks are becoming portals to not only search the web but to curate it with other people using feedback mechanisms

(37Signals.com, August 21, 2007)
Contextual advertising for an ad for a dog walker, strategically placed in a location that a dog walker may often see.


(Designer Daily, August 8, 2007)
Another example of contextual advertising

Contextual advertising only focuses on keywords and phrases in the text of a website. While many words have multiple meanings, simply placing an advertisement on a web page that is linked to a certain word can often be a problem. Take for instance the examples below.


(chiefmartec.com, December 13, 2008)


(chiefmartec.com, December 13, 2008)

Computer systems chose the placement of these ads based on keywords instead of humans understanding the meaning of the information to make an informed decision about what type of ad should be shown. This is where semantic targeting comes in.


(Semantic Spectrum, mkberkman.com)

Semantic targeting analyzes every web page in order to understand the intended meaning so the system can place the most appropriate ad on each page. It identifies the senses of those words all together. Semantic advertising using ontologies, which is a representation of a set of concepts and the relationships between them. Google claims that it is "a family of ontologies that describe the semantic web." Semantic advertising also uses taxonomies, which are classifications of organisms, along with natural learning processing and machine learning.


(Stefan Decker on the Semantic Web)

Websites such as Facebook use demographics and psychographics. Facebook's definition of psychographics is "the study and classification of people according to their attitudes, aspirations, and other psychological criteria".

The negative side of contextual, behavioral, and semantic targeting, is that some users feel like it is a breach of their privacy. One must ask where the limits are to tracking online behavior, and how much we do to ourselves by putting information online and surfing the web?