Lecture 1
Algorithms:
Set of rules to obtain the expected output from the given input
Encoded procedure for transforming input data into a desired output, based on specific
calculations
Types of algorithms (related to media):
- Prioritisation (making an ordered list)
- Classification (picking a category)
- Association (finding links)
- Filtering (isolating what’s important)
Rule based algorithms:
Based on a set of rules or steps
IF ‘condition x’ THEN ‘result y’
Quick and easy to follow but is just for specific conditions
Machine learning algorithms:
Algorithms that ‘learn’ by themselves (based on statistical models)
‘Trained’ based on a data from which they may learn to make certain kinds of decision
without human oversight
Flexible and amenable to adaptation but need to be trained
Recommender systems:
Algorithms that provide suggestions for content that is more likely to be of interest to a
particular user
Users receive distinctive streams of online content
To avoid choice overload , to maximise user relevance and to increase work efficiency
Techniques:
- Content based filtering - recommendations similar to the ones the user liked in the
past
- Collaborative filtering - recommendations based on the items that other users with
similar taste liked in the past
- Hybrid filtering - algorithms combine features from both content-based and
collaborative systems, and usually with other additional elements
Algorithm appreciation:
People rely more on advice from algorithms than from other people (despite blindness to
algorithms process)
Automation bias → humans tend to over-rely on automation because computers are rational
Algorithmic aversion:
Tendency to prefer human judgement over algorithm decisions
Less tolerance for errors from algorithms than from humans
People adverse because they don’t tend to understand the algorithmic process →
algorithmic anxiety (lack of control and uncertainty over algorithm)
, Aversion and appreciation depend on:
- Type of task
- Level of subjectivity in decisions
- Individual characteristics
Lecture 2
Online behavioural advertising → The practice of monitoring people’s online behaviour and
using the collected information to show people individually targeted advertisements
Factors:
Advertiser controlled -
Ad characteristics (level of personalisation and accuracy)
OBA transparency
Consumer controlled -
Knowledge and abilities (consumers tend to have little knowledge on OBA and legal
protections)
Consumer perceptions (mixed - OBA is relevant but also creepy)
Consumer characteristics (depends on levels of privacy concerns how you respond to OBA)
OBA outcomes:
Advertising effects are nuanced (personalisation paradox - you perceive something as
relevant and thus click on it but you still have feelings of vulnerability)
Personalisation paradox → personalised ads lead to more purchase intention but also more
scepticism and therefore to less purchase intention
Online psychological persuasion:
Personalised advertising based on the type of people you are
Big 5 → plot people in one of the five dimensions (introvert to extrovert) in regards to their
personality traits
Study 1:
Extraversion → personality trait reflecting the extent to which people seek and enjoy
company, excitement and stimulation
Users were more likely to purchase the product after viewing an ad that matched their
personality
Study 2:
Openness → personality trait reflecting the extent to which people prefer novelty over
convention
Users where more likely to install the app after viewing an as that matches their personality
(for low openness audience)
Study 3:
Algorithms:
Set of rules to obtain the expected output from the given input
Encoded procedure for transforming input data into a desired output, based on specific
calculations
Types of algorithms (related to media):
- Prioritisation (making an ordered list)
- Classification (picking a category)
- Association (finding links)
- Filtering (isolating what’s important)
Rule based algorithms:
Based on a set of rules or steps
IF ‘condition x’ THEN ‘result y’
Quick and easy to follow but is just for specific conditions
Machine learning algorithms:
Algorithms that ‘learn’ by themselves (based on statistical models)
‘Trained’ based on a data from which they may learn to make certain kinds of decision
without human oversight
Flexible and amenable to adaptation but need to be trained
Recommender systems:
Algorithms that provide suggestions for content that is more likely to be of interest to a
particular user
Users receive distinctive streams of online content
To avoid choice overload , to maximise user relevance and to increase work efficiency
Techniques:
- Content based filtering - recommendations similar to the ones the user liked in the
past
- Collaborative filtering - recommendations based on the items that other users with
similar taste liked in the past
- Hybrid filtering - algorithms combine features from both content-based and
collaborative systems, and usually with other additional elements
Algorithm appreciation:
People rely more on advice from algorithms than from other people (despite blindness to
algorithms process)
Automation bias → humans tend to over-rely on automation because computers are rational
Algorithmic aversion:
Tendency to prefer human judgement over algorithm decisions
Less tolerance for errors from algorithms than from humans
People adverse because they don’t tend to understand the algorithmic process →
algorithmic anxiety (lack of control and uncertainty over algorithm)
, Aversion and appreciation depend on:
- Type of task
- Level of subjectivity in decisions
- Individual characteristics
Lecture 2
Online behavioural advertising → The practice of monitoring people’s online behaviour and
using the collected information to show people individually targeted advertisements
Factors:
Advertiser controlled -
Ad characteristics (level of personalisation and accuracy)
OBA transparency
Consumer controlled -
Knowledge and abilities (consumers tend to have little knowledge on OBA and legal
protections)
Consumer perceptions (mixed - OBA is relevant but also creepy)
Consumer characteristics (depends on levels of privacy concerns how you respond to OBA)
OBA outcomes:
Advertising effects are nuanced (personalisation paradox - you perceive something as
relevant and thus click on it but you still have feelings of vulnerability)
Personalisation paradox → personalised ads lead to more purchase intention but also more
scepticism and therefore to less purchase intention
Online psychological persuasion:
Personalised advertising based on the type of people you are
Big 5 → plot people in one of the five dimensions (introvert to extrovert) in regards to their
personality traits
Study 1:
Extraversion → personality trait reflecting the extent to which people seek and enjoy
company, excitement and stimulation
Users were more likely to purchase the product after viewing an ad that matched their
personality
Study 2:
Openness → personality trait reflecting the extent to which people prefer novelty over
convention
Users where more likely to install the app after viewing an as that matches their personality
(for low openness audience)
Study 3:
