Marketing analytics alpesh doshi social network analysis - using social graph constructs

Using Social Graph constructs to understand user behaviour, recommendations and influence

Marketing Analytics Conference

Alpesh Doshi, Fintricity

Social Network Analysis Tuesday 6th December 2011

• What is a Social Graph? Social Network Analysis?

• Examples of Graphs

• Characteristics of Social Graphs?

• How can you use Graphs in the real world?

• The Future

• Summary

Social Graphs and their application

Agenda

What is a Social Graph? Social Network Analysis

The definitions of these set the foundation of a network of social objects a foundation for a new business approach

• Social Network Analysis (SNA) has its origins in both social science and network theory and graph theory – Network Theory concerns itself with the formulation and solution

of problems that have a network structure – usually captured in a graph

– Graph Theory provides a set of abstract concepts and methods for the analysis of graphs

• Social Graph - The social graph is a term coined by scientists working in the social areas of graph theory. It has been described as "the global mapping of everybody and how they're related". The term was popularized at the Facebook

What is a Social Graph?

A Social graph can be extremely complex and models relationships between – either inside or outside the enteprise – OR both!

• People are connected to each other directly and indirectly.

• The way they are connected varies. It could be through work, friendship, common interests, etc.

• More importantly, there is something common between the people that are connected. An implicity or explicity connection

• Connections are spread across the social web – not only on one site or application

What is a Social Graph?

People, connected. Or further, ‘social objects’ connected.

What could you capture on a Social Graph? Capturing more than just a relationship can make a graph more useful

Demographics Age, Gender,

Geography, HHI, Level of Education,

List of friends, Friends of Friends

Interests Profile-Based, Contextual,

Demonstrated, Undeclared

Actions Creating, Rating, Sending,

Sharing, Uploading, Watching, and more

Interaction How people interact with

content and ads: Clicks, time spent, interactions, videos

completed

Recency and Frequency How often and when

people express interests or actions

Sentiment and Exposure What people say, what

they read, and when and how they say and read it

• Strong Ties and Weak Ties

• Centrality – Defining people who are highly connected

• Degree – How many people can this person reach directly?

• Betweenness – How likely is this person to be the most direct route between two people in the network?

• Closeness – How fast can this person reach everyone in the network?

• EigenVector – How is this person connected to well connected people?

Some Characteristics of a graph Multi-relational graphs are the foundation. These are graphs where relationships are described in more than one way.

• Recommendation Engines

• Interest Graphs

• Influence Networks

• Sentiment Analysis

• Searching, Scoring, Ranking

Where can you use Graphs? How can you apply social graphs, and social network analysis for business benefit? Some examples

• Recommendation engines need a ‘boost’ – and can apply to many different types of recommendation • Recommend to read articles/content

• Recommend collaborators

• Recommend sites to post content

• Recommend people with similar interests

Where can you use Graphs? Recommendation Recommendation engines are a natural fit for graphs.

• Interest Graphs (with Linked Data) • Represents your interests

• Represents your interest in relation to others(social graph)

• Graphs can weaken/strengthen your interests dependent on behavious

• Recommend people with similar interests

Where can you use Graphs? Recommendation Recommendation engines are a natural fit for graphs.

• First generation influence tools have shown that there is a demand (e.g. peer index, klout)

• Useful only in content (e.g. Oil/Gas, Entertainment)

• A sub-set of a Social graph

• Influence drives many things recommendation, opinion, purchasing

Where can you use Graphs? Influence Social Media is driving the use of influence – PR Agencies, enterprise have to now can understand influence in detail

• Searching • Combining a social graph with linked data (semantics and ontologies)

• Graphs can be searched in different ways to discover relevant information

• Scoring • Score individuals, content, objects to create patterns of associations

• Can be used for targeted advertising, content selection

• Ranking

• Ranking use, importance, connectedness etc

• Marketing uses included finding the most important customers (together with modelling influence

Where can you use Graphs? Searching, Scoring, Ranking This can help identify and segment users by different categories together with the social graph

• Social Graphs are not new!

• But their uses in marketing are nascent

• The implementation of a scalable graph database technology is relatively new – therefore can better implement solutions around this space.

• Application of social graphs to marketing allow more targeted, integrated and cost effective solutions

Conclusions Social Graphs will change the way marketing is done in the future. It will become a fabric and key component in any marketing activity

Alpesh Doshi, Fintricity e: [email protected]

o: +44 870 020 1656, m: +44 7973 822820

Linkedin: www.linkedin.com/in/alpeshdoshi

Twitter: @alpeshdoshi

Questions?

Content from this presentation taken from Giorgos Cheliotis under Creative Commons