Automated semiotic analysis uses machine learning techniques to analyze large amounts of text and imagery to quantify their underlying semiotic structures. This expands the traditional role of semiotics, which has typically been done by human experts at a small scale. The webinar discusses how automation can help semioticians by enabling them to study far more data than previously possible. It provides examples of using automated analysis to study images and social media posts from various brands to identify patterns in archetypes, values, and universal energies to gain new cultural and brand insights. Semiotic analysis can now be applied at a much larger scale to problems like analyzing websites, marketing material, and media.
3. This is NOT about replacing semioticians
It’s about EXPANDING the role of semiotics.
It’s about enabling semioticians.
4. Hero, Heretic or Helper
Hero – does semiotics need heroics?
Heretic – is semiotics an approach that just does not need computers/AI/ML etc?
Helper – is automation going to help semiotics?
5. Charles Sanders Peirce.
“Charles Sanders Peirce: A
Life
By Joseph Brent.”
“Wasp in a bottle”
“
William James.
“Principles of Psychology”
1890.
6.
7. Semiotics is the Art and Science of ‘signs’ – knowledge
structures of text and imagery. It’s traditionally been
done by human experts at small scale.
Quantitative Semiotics uses machine learning
techniques to analyse text and imagery, to quantify
the semiotic knowledge structures.
11. Semiotics tend to be practitioner based: individuals.
This leads to a scaling problem, humans can only consistently analyse a limit amount
of information.
Automated analysis is not a total solution to the problem of scale but it is a start.
New approaches need to be developed to deal with the scale of information
available.
12. 12,000
Instagram
posts from
20 brands
Appletiser
Coca Cola
Dirty Lemon
Tango
Dr Pepper
Evian
Fever Tree
Innocent
Irnbru
Karma Drinks
Lipton Ice Tea
Lucozade Energy
Lucozade Sport
Orangina
Pepsi
Perrier
Real Kombucha
Red Bull
Sandows
Tropicana Juices
DATA:
Post image
Post text
Date of post
Number of likes
Number of comments
Follower count
14. 1
2
3
4
5
Universalism
Benevolence
Conformity
Traditionalism
Security
Power
Achievement
Hedonism
Stimulation
Self directedness
Coca Cola
Innocent
Lucozade Sport
Schwartz’s work defined a range of cross-cultural
values that relate to individuals’ worldview and
value systems
They can also provide insight into what brands stand
for in culture based on what they are communicating.
• Universalism – fairness, understanding, tolerance, and protection
in society and for nature.
• Benevolence – preserving and enhancing the welfare of close
individuals and the community.
• Tradition – respect, commitment, and acceptance of the customs
and ideas that are important in culture.
• Conformity – restraint of actions, inclinations, and impulses likely
to upset or harm others and violate social expectations or norms.
• Security – safety, harmony, and stability of society, of
relationships, and of self.
• Power – social status and prestige, control or dominance over
people and resources.
• Achievement – success through achieving personal and social
goals and desires.
• Hedonism – indulgence, pleasure or sensuous gratification.
• Stimulation – excitement, novelty, and challenge in life.
• Self-Direction – independent thought and action; choosing,
creating, exploring.
Archetypes: “narrative character” or “thematic structure”
20. Automated semiotic analysis allows the analysis of much larger data sets than human practitioners can
manage.
Automated semiotic analysis is a tool that expands the use of semiotic analysis.
Semiotics is no longer an “addition” within a project but becomes the core analysis technique for a project.
The ability for semiotic analysis to “scale up” means it can be used on a wider class of problems:
Semiotic analysis of web sites
Semiotic analysis of sales and marketing material
”Cultural Weather” analysis – internal/external communication analysis
”General” cultural analysis – media et al
Implications