SlideShare a Scribd company logo

Social Media Analytics Department For Work and Pensions Research Seminar

Download as PPTX, PDF
Dr Wasim Ahmed, profile picture
Dr Wasim Ahmed

The document presents a seminar by Wasim Ahmed on social media analytics, particularly focusing on Twitter's role in public communication and data analysis. It covers the popularity and tools for real-time social media analysis, discussing various case studies and analytics methods, including sentiment and network analysis. Additionally, it highlights software like Visibrain and Echosec used for monitoring and analyzing social media data.

Social Media
Related topics:
Data Mining Insights
1 of 67
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
Social Media Analytics Wasim Ahmed (wahmed1@Sheffield.ac.uk) (Twitter: @was3210) Department for Work and Pensions Social Media Research Seminars. 4th August 2016, London.
05/08/2016 ยฉ The University of Sheffield 2 About me โ€ข Second Year PhD student in the Health Informatics Research Group, Information School, University of Sheffield. โ€ข Research Associate (Social Media Specialist), Management School, University of Sheffield. โ€ข Visibrain Blog Partner, Connected Action Partner, Echosec Advisor, SMRF Member, and NSMNSS member.
05/08/2016 ยฉ The University of Sheffield 3 PhD Background โ€ข PhD project examines infectious disease outbreaks on Twitter (Ebola and Swine Flu). โ€ข PhD contains a chapter on tools, methods and software of capturing and analysing Twitter data. โ€ข Published articles on social media software on my research blog, on the LSE Impact blog, and for the MmIT journal.
05/08/2016 ยฉ The University of Sheffield 4 Overview โ€ข Part 1 โ€“ Outlines Twitter as a platform, and the popularity of Twitter research, and some methods of analysis. โ€ข Part 2 โ€“ Examines tools available for real time social media analysis.
05/08/2016 ยฉ The University of Sheffield 5 Part 1 Part 1 of the seminar will outline: โ€ข Twitter as a platform โ€ข Popularity of Twitter โ€ข Brand Awareness โ€ข Methods of analysis
05/08/2016 ยฉ The University of Sheffield 6 Twitter โ€ข Twitter allows brief <140 character text updates, know as โ€˜tweetsโ€™, to be shared with other users as well as images, videos, and links. โ€ข Tweets can contain thoughts, feelings, activities, and opinions (Chew and Eysenbach, 2010).
05/08/2016 ยฉ The University of Sheffield 7 Popularity of Twitter โ€ข Twitter is the most popular platform in regards to academic research and for companies producing social media analytics. โ€ข Why is Twitter so popular?
05/08/2016 ยฉ The University of Sheffield 8 Popularity of Twitter โ€ข Twitter is a popular platform in terms of the media attention it receives and it therefore attracts more research due to its cultural status. โ€ข Twitter makes it easier to find and follow conversations (i.e., by both its search feature and by tweets appearing in Google search results).
05/08/2016 ยฉ The University of Sheffield 9 Popularity of Twitter โ€ข Twitter has hashtags which make it easier gathering, sorting, and expanding searches when collecting data. โ€ข Therefore Twitter data is easy to retrieve as major incidents, news stories and events on Twitter tend to be centred around a hashtag. e.g..#Brexit
05/08/2016 ยฉ The University of Sheffield 10 Popularity of Twitter โ€ข The Twitter API is more open and accessible compared to other social media platforms, which makes Twitter more favourable to developers creating tools to access data. This consequently increases the availability of tools to researchers. โ€ข Many researchers themselves are using Twitter and because of their favourable personal experiences, they feel more comfortable with researching a familiar platform.
05/08/2016 ยฉ The University of Sheffield 11 Why are these applications important? โ€ข When it became clear that Twitter was becoming an important social networking site and public communication platform a number of businesses and social media marketing professionals attempted to exploit the platform for commercial purposes
Toyota โ€ข Toyota had to recall a number of its cars in 2009 and 2010 due to a serious safety fault which resulted in the deaths of over 50 people 05/08/2016 ยฉ The University of Sheffield 12
โ€ข As soon as the recall crisis start getting media attention Toyota quickly put together an โ€˜Online Newsroomโ€™ and a โ€˜Social Media Strategy Teamโ€™ to coordinate all the media releases 05/08/2016 ยฉ The University of Sheffield 13 Toyota
Sony PlayStation Network โ€ข In mid-April 2011 the Playstation Network was suddenly shut down without explanation โ€ข Frustrations quickly spread through social media sites such as Twitter as gamers around the world voiced their annoyance at not being able to access their online games 05/08/2016 ยฉ The University of Sheffield 14
Sony PlayStation Network โ€ข The lack of regular updates and information from Sony served to incense users โ€ข Users struggled to determine what was fact and what was rumour on Twitter 05/08/2016 ยฉ The University of Sheffield 15
โ€ข โ€œI think It is pretty disgusting that Sony have waiting 7 days to tell users that their Credit Card details may have been compromisedโ€. โ€ข โ€œI bet the hacker will get emails out quicker than Sony!โ€ 05/08/2016 ยฉ The University of Sheffield 16 Sony PlayStation Network
Toyota โ€ข While there was still anger and negative viewpoints shared through social media services, โ€ข Company was able to minimise their impact by eliminating confusion and keeping the consumer base regularly informed of developments 05/08/2016 ยฉ The University of Sheffield 17
Brand Management โ€ข The two cases have highlighted brands need to know how they are being mentioned across social media profiles โ€ข Social Media Analytics is now a huge market 05/08/2016 ยฉ The University of Sheffield 18
05/08/2016 ยฉ The University of Sheffield 19 API โ€ข An API in simple terms is a set of requirements which dictate how one application may talk to another โ€ข On the web APIs allow services such as Google maps or Facebook let other application โ€˜piggybackโ€™ on their services โ€ข APIs allow outside programs to access specific sets of features thus saving developers time and resources i.e., accessing data See http://readwrite.com/2013/09/19/api-defined/
05/08/2016 ยฉ The University of Sheffield 20 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข i) Twitterโ€™s Search API โ€“ similar to the search function in Twitter. It is focused on relevance and not completeness, some tweets and users may be missing from results (possible to go back in time 7 days from the query date up to 3200 queries)
05/08/2016 ยฉ The University of Sheffield 21 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข ii) Twitterโ€™s Streaming API โ€“ Streaming APIs give developers low latency access to Twitterโ€™s global stream of tweet data (a type of โ€˜liveโ€™ stream)
05/08/2016 ยฉ The University of Sheffield 22 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข iii) Firehose API โ€“ in theory, 100% of Twitter data (most software allows up to 30 days worth of historical tweets)
05/08/2016 ยฉ The University of Sheffield 23 Legality โ€ข Sharing Twitter datasets is prohibited under Twitterโ€™s terms and conditions see: https://dev.twitter.com/overview/terms/agreeme nt-and-policy โ€ข However, sharing Tweet IDs (to look up the tweets that are used is permissible).
05/08/2016 ยฉ The University of Sheffield 24 Examples of Methods Used to Analyse Twitter Data โ€ข Sentiment analysis has the potential to work well with Twitter data, as tweets are short (i.e., <=140 characters) โ€ข However, sarcasm is difficult to detect within tweets. โ€ข See the SentiStrength algorithm (http://sentistrength.wlv.ac.uk/)
05/08/2016 ยฉ The University of Sheffield 25 Types of Analysis โ€ข Time series analysis can be used to examine tweets overtime and to see when a peak in tweets may occur or have occurred.
05/08/2016 ยฉ The University of Sheffield 26 See Example of DWP
Types of Analysis โ€ข Network analysis is used to visualize the connections between users and visually display the connections โ€ข Network analysis can also be used to cluster Twitter users into different groups 05/08/2016 ยฉ The University of Sheffield 27
Types of Analysis โ€ข Machine Learning โ€ข Example: Manually coding a subset of data e.g., 100 tweets in a dataset of a 1,000 where the algorithm will automatically classify the remaining data. 05/08/2016 ยฉ The University of Sheffield 28
Types of Analysis โ€ข Content Analysis/Thematic Analysis โ€ข Example: Manually placing tweets into predefined categories and/or exploring tweets and developing categories. โ€ข Use Microsoft Excel/NVivo 05/08/2016 ยฉ The University of Sheffield 29
Part 2 โ€ข Part 2 of the seminar will provide an overview of some of the most popular social media analysis and analytical software available to the public. 05/08/2016 ยฉ The University of Sheffield 30
Visibrain 05/08/2016 ยฉ The University of Sheffield 31
Visibrain โ€ข Visibrain is a powerful media monitoring tool which has access to the Twitter Firehose โ€ข Therefore, it is not limited in the amount that can be retrieved unlike the Search and Stream APIs. 05/08/2016 ยฉ The University of Sheffield 32
Visibrain โ€ข Explore and filter tweets, links, users, hashtags, expressions, mentions, countries, languages, genders, audiences, occupations, interests, client types, content types, and domains. 05/08/2016 ยฉ The University of Sheffield 33
Possible uses โ€ข Security teams, organisations, and/or government โ€ข For example you could monitor Twitter/social media for mention of a bank and a hacking group and/or a city and โ€˜attackโ€™ โ€ข If triggered you would receive an alert 05/08/2016 ยฉ The University of Sheffield 34
Visibrain 05/08/2016 ยฉ The University of Sheffield 35
Visibrain โ€ข Receive reports produced and delivered almost instantly displaying tweet content, actors, URLS, and so forth. โ€ข Possible to receive these alerts via email to almost any location in the world. 05/08/2016 ยฉ The University of Sheffield 36
Visibrain 05/08/2016 ยฉ The University of Sheffield 37
Echosec 05/08/2016 ยฉ The University of Sheffield 38
Echosec โ€ข Many citizens now carry mobile devices connected to the Internet capable of taking video and pictures โ€ข Emerging news stories of crisis events may be reported on social media before reported anywhere else. These posts may contain location information. 05/08/2016 ยฉ The University of Sheffield 39
Echosec โ€ข For example, those whom are affected by a flood may post information and media with location information such as video or pictures to social media which can then be utilized by the emergency services 05/08/2016 ยฉ The University of Sheffield 40
Echosec โ€ข Echosec has removed the difficulty of identifying relevant content by providing the ability to search for locations as well as content 05/08/2016 ยฉ The University of Sheffield 41
Echosec โ€ข Echosec allows you to navigate to almost any location in the world and examine the social media activity around that area โ€ข Currently Echosec Pro allows users to access at least the following social data feeds: Instagram, Twitter, Foursquare, AIS Shipping, Sina Weibo, Flickr, YouTube, VK, and Panoramio 05/08/2016 ยฉ The University of Sheffield 42
Possible uses of Echosec โ€ข Crisis events: โ€ข Floods โ€ข Earthquakes โ€ข Situational awareness: โ€ข Concerts โ€ข Sports events โ€ข Security Teams: โ€ข Government security teams/ Local Authorities โ€ข Private Investigators 05/08/2016 ยฉ The University of Sheffield 43
Possible uses of Echosec 05/08/2016 ยฉ The University of Sheffield 44 โ€ข Geographically searching social media data in an area can show you what you are looking for in an emergency โ€ข Can examine locations of affected areas and see where people have stopped posting from
Possible Use Case 05/08/2016 ยฉ The University of Sheffield 45 โ€ข Security services e.g., NSA examining content shared from building premises for security breaches โ€ข Echosec has found cases of employees working within secure buildings to be taking selfies with their ID badges
Dashboard layout of Echosec 05/08/2016 ยฉ The University of Sheffield 46
NodeXL 05/08/2016 ยฉ The University of Sheffield 47
NodeXL โ€ข Social media analysis that looks at the structure of networks โ€ข Produces analytics such as most frequently occurring URLs, domains, hashtags, keywords, overall and by group level. 05/08/2016 ยฉ The University of Sheffield 48
NodeXL โ€ข Social media analysis that looks at the structure of networks โ€ข Produces analytics such as most frequently occurring URLs, domains, hashtags, keywords, overall and by group level. 05/08/2016 ยฉ The University of Sheffield 49
NodeXL โ€ข Essentially NodeXL allows you to take pictures of crowds โ€ข Graph Gallery โ€“ a โ€œFlickr for network graphsโ€ โ€“ Marc Smith (Social Media research foundation) โ€ข DWP network graph by Connected Action Consulting Group 05/08/2016 ยฉ The University of Sheffield 50
NodeXL 05/08/2016 ยฉ The University of Sheffield 51
NodeXL โ€“ 6 types of network structures 05/08/2016 ยฉ The University of Sheffield 52
NodeXL โ€“ 6 types of network structures 05/08/2016 ยฉ The University of Sheffield 53
05/08/2016 ยฉ The University of Sheffield 54 NodeXL
05/08/2016 ยฉ The University of Sheffield 55 NodeXL interactive graph
05/08/2016 ยฉ The University of Sheffield 56 NodeXL G1 MP for Ashford. Secretary of State for Work and Pensions.
05/08/2016 ยฉ The University of Sheffield 57 NodeXL G2 DWP Twitter account
NodeXL โ€ข [640] Words in Sentiment List#1: Positive โ€ข [1544] Words in Sentiment List#2: Negative โ€ข [4] Words in Sentiment List#3: Angry/Violent 05/08/2016 ยฉ The University of Sheffield 58
NodeXL โ€ข Top Word Pairs in Tweet in Entire Graph: [325] phone,lines [319] lines,dwp [214] scrap,rip [196] rip,phone [169] expensive,phone [169] ministry,labour [159] calls,dwp [139] scrap,dwp [137] dwp,minister [128] new,dwp 05/08/2016 ยฉ The University of Sheffield 59
NodeXL โ€ข [160] Social Security Advisory Committee, Telephony in DWP and HMRC: an update https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/5 38834/telephony-in-dwp-and-hmrc-an-update-op17.pdf โ€ข [68] Benefit sanctions handed out to thousands of people โ€˜unlawfulโ€™ http://www.independent.co.uk/news/uk/politics/dwp-defeated-in-court-again- over-unlawful-retroactive-benefit-sanctions-a7006466.html โ€ข [32] Make all calls to the department for work and pensions free https://you.38degrees.org.uk/petitions/make-ian-duncan-smith-make-all-phone- communication-free?bucket=blast&source=twitter-share-button โ€ข [27] Owen Smith vows to scrap the DWP as โ€˜cold-eyed revolutionโ€™ in Labour leadership speech http://www.mirror.co.uk/news/uk-news/owen-smith- labour-leadership-revolution-8502537 05/08/2016 ยฉ The University of Sheffield 60
NodeXL 05/08/2016 ยฉ The University of Sheffield 61 โ€ข Although the graph gallery contains a number of network graphs from Twitter - it is possible to import networks from other social media platforms such as Facebook and Instagram
DiscoverText 05/08/2016 ยฉ The University of Sheffield 62
05/08/2016 ยฉ The University of Sheffield 63 DiscoverText โ€ข Useful for machine learning, de- duplication, and collaboration. โ€ข Used with Sifter, it is possible to retrieve data going back in time more than 30 days โ€ข Data cost will range from not that expensive to very expensive depending on the query and time
05/08/2016 ยฉ The University of Sheffield 64 DiscoverText โ€ข University of Sheffield Project โ€ข Analysed the mentions of the University over a 1 year period and classified tweets โ€ข Better understand who engages with the University on Twitter
05/08/2016 ยฉ The University of Sheffield 65 Conclusion โ€ข There is no best social media analytics or analysis software โ€“ each offers something different. โ€ข I tend to use a number of software in combination to gain the best insight. โ€ข Thanks for listening.
05/08/2016 ยฉ The University of Sheffield 66 Questions?
To Discover And Understand.

More Related Content

PPTX
An Introduction to NodeXL
Dr Wasim Ahmed
ย 
PPTX
Life after High Storrs: PhD study and social media research
Dr Wasim Ahmed
ย 
PPTX
Introduction to software that can be used to capture and analyse Twitter data
Dr Wasim Ahmed
ย 
PPTX
Social Media for Marketing An Overview of Specialist Software
Dr Wasim Ahmed
ย 
PPTX
An overview of Twitter analytics
Dr Wasim Ahmed
ย 
PPTX
Communicating Science Through Social Media: Tools and Techniques
Dr Wasim Ahmed
ย 
PPTX
Social Media: A Practical Approach
Dr Wasim Ahmed
ย 
PPTX
Nordmedia 2013 Villi, Matikainen & Khaldarova
jannem
ย 
An Introduction to NodeXL
Dr Wasim Ahmed
ย 
Life after High Storrs: PhD study and social media research
Dr Wasim Ahmed
ย 
Introduction to software that can be used to capture and analyse Twitter data
Dr Wasim Ahmed
ย 
Social Media for Marketing An Overview of Specialist Software
Dr Wasim Ahmed
ย 
An overview of Twitter analytics
Dr Wasim Ahmed
ย 
Communicating Science Through Social Media: Tools and Techniques
Dr Wasim Ahmed
ย 
Social Media: A Practical Approach
Dr Wasim Ahmed
ย 
Nordmedia 2013 Villi, Matikainen & Khaldarova
jannem
ย 

What's hot (20)

PDF
Practical Tools Social Media For Consumer Insight (Guest Lecture)
Dr Wasim Ahmed
ย 
PDF
Building Student Digital Capability In Computing And Digital Technologies Thr...
Thomas Lancaster
ย 
PDF
Examining The State of Academic Integrity in Europe - University Politehnica ...
Thomas Lancaster
ย 
PDF
The State Of Academic Integrity In Europe - Taras Shevchenko University, Kiev...
Thomas Lancaster
ย 
PDF
Scientists go online: Scientific social networking and Open Notebook Science
Lucy Power
ย 
PPTX
Open citations: Next steps
Ludo Waltman
ย 
PPT
Electronic resources and student attainment - Phil Adams
Eduserv
ย 
PDF
Lessons Learnt from LinkedUp
Marieke Guy
ย 
PDF
Open Education Challenge 2014: exploiting Linked Data in Educational Applicat...
Stefan Dietze
ย 
PPTX
Responsible journals: Making reading, evaluation and publishing open
Ludo Waltman
ย 
PDF
Research support and open science services at the University of Eastern Finla...
Library_Connect
ย 
PPTX
Crossref LIVE19 - Researcher and metadata user view
Ludo Waltman
ย 
PPTX
Introducing the IRUSdataUK pilot - Jisc Digifest 2016
Jisc
ย 
PDF
Reincarnation App - Prototype
Sylvia van Schie
ย 
PPT
Demaeyer
Knight Center
ย 
PPTX
Large-scale visualization of science: Methods, tools, and applications
Ludo Waltman
ย 
PDF
International Perspectives On Contract Cheating - Staffordshire University St...
Thomas Lancaster
ย 
PDF
Why do teachers (not) use the institutional repository?
Juliร  Minguillรณn
ย 
PPTX
Visualizing science using VOSviewer based on Crossref, Microsoft Academic, an...
Nees Jan van Eck
ย 
PPTX
Visualizing science based on open data sources
Nees Jan van Eck
ย 
Practical Tools Social Media For Consumer Insight (Guest Lecture)
Dr Wasim Ahmed
ย 
Building Student Digital Capability In Computing And Digital Technologies Thr...
Thomas Lancaster
ย 
Examining The State of Academic Integrity in Europe - University Politehnica ...
Thomas Lancaster
ย 
The State Of Academic Integrity In Europe - Taras Shevchenko University, Kiev...
Thomas Lancaster
ย 
Scientists go online: Scientific social networking and Open Notebook Science
Lucy Power
ย 
Open citations: Next steps
Ludo Waltman
ย 
Electronic resources and student attainment - Phil Adams
Eduserv
ย 
Lessons Learnt from LinkedUp
Marieke Guy
ย 
Open Education Challenge 2014: exploiting Linked Data in Educational Applicat...
Stefan Dietze
ย 
Responsible journals: Making reading, evaluation and publishing open
Ludo Waltman
ย 
Research support and open science services at the University of Eastern Finla...
Library_Connect
ย 
Crossref LIVE19 - Researcher and metadata user view
Ludo Waltman
ย 
Introducing the IRUSdataUK pilot - Jisc Digifest 2016
Jisc
ย 
Reincarnation App - Prototype
Sylvia van Schie
ย 
Demaeyer
Knight Center
ย 
Large-scale visualization of science: Methods, tools, and applications
Ludo Waltman
ย 
International Perspectives On Contract Cheating - Staffordshire University St...
Thomas Lancaster
ย 
Why do teachers (not) use the institutional repository?
Juliร  Minguillรณn
ย 
Visualizing science using VOSviewer based on Crossref, Microsoft Academic, an...
Nees Jan van Eck
ย 
Visualizing science based on open data sources
Nees Jan van Eck
ย 
Ad

Similar to Social Media Analytics Department For Work and Pensions Research Seminar (20)

PPTX
Social Media Analytics Lecture
Dr Wasim Ahmed
ย 
PPTX
Keynote Talk - Gaining Powerful Insights into Social Media Listening
Dr Wasim Ahmed
ย 
PPTX
Insights From Social Media
Dr Wasim Ahmed
ย 
PPTX
Researching Social Media โ€“ Big Data and Social Media Analysis
Farida Vis
ย 
PPT
The evolution of research on social media
Farida Vis
ย 
PPTX
Social Media Analytics
Muhammad Rifqi
ย 
PPTX
Social Media Data Collection & Analysis
Scott Sanders
ย 
PPTX
Analyzing Social Media with Digital Methods. Possibilities, Requirements, and...
Bernhard Rieder
ย 
PDF
IRJET - Social Media Intelligence Tools
IRJET Journal
ย 
PDF
Sentiment Analysis and Social Media: How and Why
Davide Feltoni Gurini
ย 
PPTX
Bd & privacy issues in social media
Jorge G. Moreno Lรณpez
ย 
PPTX
Social Media Mining and Analytics
Web Science Research Group at Institute of Business Administration, Karachi, Pakistan
ย 
PPTX
Social Network Analysis with NodeXL Part 1
Dr Wasim Ahmed
ย 
PDF
People as sensors - mining social media for meaningful information
Tom Raftery
ย 
PDF
s00146-014-0549-4.pdf
EngrAliSarfrazSiddiq
ย 
PPT
Picturing the Social: Talk for Transforming Digital Methods Winter School
Farida Vis
ย 
PDF
Weller pleasures+perils social media
Katrin Weller
ย 
PDF
Introduction to the Responsible Use of Social Media Monitoring and SOCMINT Tools
Mike Kujawski
ย 
PPTX
Social Media Data Mining
Ryan Reede
ย 
PPTX
SKOPOS Defining Social Media
skoposuk
ย 
Social Media Analytics Lecture
Dr Wasim Ahmed
ย 
Keynote Talk - Gaining Powerful Insights into Social Media Listening
Dr Wasim Ahmed
ย 
Insights From Social Media
Dr Wasim Ahmed
ย 
Researching Social Media โ€“ Big Data and Social Media Analysis
Farida Vis
ย 
The evolution of research on social media
Farida Vis
ย 
Social Media Analytics
Muhammad Rifqi
ย 
Social Media Data Collection & Analysis
Scott Sanders
ย 
Analyzing Social Media with Digital Methods. Possibilities, Requirements, and...
Bernhard Rieder
ย 
IRJET - Social Media Intelligence Tools
IRJET Journal
ย 
Sentiment Analysis and Social Media: How and Why
Davide Feltoni Gurini
ย 
Bd & privacy issues in social media
Jorge G. Moreno Lรณpez
ย 
Social Media Mining and Analytics
Web Science Research Group at Institute of Business Administration, Karachi, Pakistan
ย 
Social Network Analysis with NodeXL Part 1
Dr Wasim Ahmed
ย 
People as sensors - mining social media for meaningful information
Tom Raftery
ย 
s00146-014-0549-4.pdf
EngrAliSarfrazSiddiq
ย 
Picturing the Social: Talk for Transforming Digital Methods Winter School
Farida Vis
ย 
Weller pleasures+perils social media
Katrin Weller
ย 
Introduction to the Responsible Use of Social Media Monitoring and SOCMINT Tools
Mike Kujawski
ย 
Social Media Data Mining
Ryan Reede
ย 
SKOPOS Defining Social Media
skoposuk
ย 
Ad

Recently uploaded (20)

PPTX
Strategies for Social Media App Enhancement
simarjeetsingh3033
ย 
PDF
StarNetCafeSB2012D3POYNagaworld2-Hotel-Casino-Phnom Entertainment
Mysisd3poycc Mysisd3poycc
ย 
PDF
The Fastest Way to Look Popular Buy Reactions Today
SocioCosmos
ย 
PDF
25K Btc Enabled Cash App Accounts โ€“ Safe, Fast, Verified.pdf
BestPlacestoBuyVerif
ย 
PDF
Buy Verified Cryptocurrency Accounts - Lori Donato's blo.pdf
Buy Verified Linkedin Accounts
ย 
PDF
THE ULTIMATE YOUTUBE SHORTS GROWTH......
gamernoorab
ย 
PDF
Climate Risk and Credit Allocation: How Banks Are Integrating Environmental R...
AJHSSR Journal
ย 
PDF
Instant Audience, Long-Term Impact Buy Real Telegram Members
SocioCosmos
ย 
PPTX
Types of Social Media Marketing for Business Success
DigitalThinkerHelp
ย 
PDF
Customer Churn Prediction in Digital Banking: A Comparative Study of Xai Tech...
AJHSSR Journal
ย 
PDF
Presence That Pays Off Activate My Social Growth
SocioCosmos
ย 
PDF
Transform Your Social Media, Grow Your Brand
karamjeet25
ย 
PDF
Your Best Post Vanished. Blame the Attention Economy
Virtual Employee Pvt. Ltd.
ย 
PDF
Mastering Social Media Marketing in 2025.pdf
Elevatemetricsdigital
ย 
PDF
Why Digital Marketing Matters in Todayโ€™s World Ask ChatGPT
gauravgangan6
ย 
PPT
memimpindegra1uejehejehdksnsjsbdkdndgggwksj
WarungApk
ย 
PDF
Subscribe This Channel Subscribe Back You
Mysisd3poycc Mysisd3poycc
ย 
PPTX
Developing lesson plan gejegkavbw gagsgf
JoanaBayoca
ย 
DOCX
Buy Goethe A1 ,B2 ,C1 certificate online without writing
wizzymillion918
ย 
PDF
TikTok Live shadow viewers_ Who watches without being counted
Ufki
ย 
Strategies for Social Media App Enhancement
simarjeetsingh3033
ย 
StarNetCafeSB2012D3POYNagaworld2-Hotel-Casino-Phnom Entertainment
Mysisd3poycc Mysisd3poycc
ย 
The Fastest Way to Look Popular Buy Reactions Today
SocioCosmos
ย 
25K Btc Enabled Cash App Accounts โ€“ Safe, Fast, Verified.pdf
BestPlacestoBuyVerif
ย 
Buy Verified Cryptocurrency Accounts - Lori Donato's blo.pdf
Buy Verified Linkedin Accounts
ย 
THE ULTIMATE YOUTUBE SHORTS GROWTH......
gamernoorab
ย 
Climate Risk and Credit Allocation: How Banks Are Integrating Environmental R...
AJHSSR Journal
ย 
Instant Audience, Long-Term Impact Buy Real Telegram Members
SocioCosmos
ย 
Types of Social Media Marketing for Business Success
DigitalThinkerHelp
ย 
Customer Churn Prediction in Digital Banking: A Comparative Study of Xai Tech...
AJHSSR Journal
ย 
Presence That Pays Off Activate My Social Growth
SocioCosmos
ย 
Transform Your Social Media, Grow Your Brand
karamjeet25
ย 
Your Best Post Vanished. Blame the Attention Economy
Virtual Employee Pvt. Ltd.
ย 
Mastering Social Media Marketing in 2025.pdf
Elevatemetricsdigital
ย 
Why Digital Marketing Matters in Todayโ€™s World Ask ChatGPT
gauravgangan6
ย 
memimpindegra1uejehejehdksnsjsbdkdndgggwksj
WarungApk
ย 
Subscribe This Channel Subscribe Back You
Mysisd3poycc Mysisd3poycc
ย 
Developing lesson plan gejegkavbw gagsgf
JoanaBayoca
ย 
Buy Goethe A1 ,B2 ,C1 certificate online without writing
wizzymillion918
ย 
TikTok Live shadow viewers_ Who watches without being counted
Ufki
ย 

Social Media Analytics Department For Work and Pensions Research Seminar

  • 1. Social Media Analytics Wasim Ahmed (wahmed1@Sheffield.ac.uk) (Twitter: @was3210) Department for Work and Pensions Social Media Research Seminars. 4th August 2016, London.
  • 2. 05/08/2016 ยฉ The University of Sheffield 2 About me โ€ข Second Year PhD student in the Health Informatics Research Group, Information School, University of Sheffield. โ€ข Research Associate (Social Media Specialist), Management School, University of Sheffield. โ€ข Visibrain Blog Partner, Connected Action Partner, Echosec Advisor, SMRF Member, and NSMNSS member.
  • 3. 05/08/2016 ยฉ The University of Sheffield 3 PhD Background โ€ข PhD project examines infectious disease outbreaks on Twitter (Ebola and Swine Flu). โ€ข PhD contains a chapter on tools, methods and software of capturing and analysing Twitter data. โ€ข Published articles on social media software on my research blog, on the LSE Impact blog, and for the MmIT journal.
  • 4. 05/08/2016 ยฉ The University of Sheffield 4 Overview โ€ข Part 1 โ€“ Outlines Twitter as a platform, and the popularity of Twitter research, and some methods of analysis. โ€ข Part 2 โ€“ Examines tools available for real time social media analysis.
  • 5. 05/08/2016 ยฉ The University of Sheffield 5 Part 1 Part 1 of the seminar will outline: โ€ข Twitter as a platform โ€ข Popularity of Twitter โ€ข Brand Awareness โ€ข Methods of analysis
  • 6. 05/08/2016 ยฉ The University of Sheffield 6 Twitter โ€ข Twitter allows brief <140 character text updates, know as โ€˜tweetsโ€™, to be shared with other users as well as images, videos, and links. โ€ข Tweets can contain thoughts, feelings, activities, and opinions (Chew and Eysenbach, 2010).
  • 7. 05/08/2016 ยฉ The University of Sheffield 7 Popularity of Twitter โ€ข Twitter is the most popular platform in regards to academic research and for companies producing social media analytics. โ€ข Why is Twitter so popular?
  • 8. 05/08/2016 ยฉ The University of Sheffield 8 Popularity of Twitter โ€ข Twitter is a popular platform in terms of the media attention it receives and it therefore attracts more research due to its cultural status. โ€ข Twitter makes it easier to find and follow conversations (i.e., by both its search feature and by tweets appearing in Google search results).
  • 9. 05/08/2016 ยฉ The University of Sheffield 9 Popularity of Twitter โ€ข Twitter has hashtags which make it easier gathering, sorting, and expanding searches when collecting data. โ€ข Therefore Twitter data is easy to retrieve as major incidents, news stories and events on Twitter tend to be centred around a hashtag. e.g..#Brexit
  • 10. 05/08/2016 ยฉ The University of Sheffield 10 Popularity of Twitter โ€ข The Twitter API is more open and accessible compared to other social media platforms, which makes Twitter more favourable to developers creating tools to access data. This consequently increases the availability of tools to researchers. โ€ข Many researchers themselves are using Twitter and because of their favourable personal experiences, they feel more comfortable with researching a familiar platform.
  • 11. 05/08/2016 ยฉ The University of Sheffield 11 Why are these applications important? โ€ข When it became clear that Twitter was becoming an important social networking site and public communication platform a number of businesses and social media marketing professionals attempted to exploit the platform for commercial purposes
  • 12. Toyota โ€ข Toyota had to recall a number of its cars in 2009 and 2010 due to a serious safety fault which resulted in the deaths of over 50 people 05/08/2016 ยฉ The University of Sheffield 12
  • 13. โ€ข As soon as the recall crisis start getting media attention Toyota quickly put together an โ€˜Online Newsroomโ€™ and a โ€˜Social Media Strategy Teamโ€™ to coordinate all the media releases 05/08/2016 ยฉ The University of Sheffield 13 Toyota
  • 14. Sony PlayStation Network โ€ข In mid-April 2011 the Playstation Network was suddenly shut down without explanation โ€ข Frustrations quickly spread through social media sites such as Twitter as gamers around the world voiced their annoyance at not being able to access their online games 05/08/2016 ยฉ The University of Sheffield 14
  • 15. Sony PlayStation Network โ€ข The lack of regular updates and information from Sony served to incense users โ€ข Users struggled to determine what was fact and what was rumour on Twitter 05/08/2016 ยฉ The University of Sheffield 15
  • 16. โ€ข โ€œI think It is pretty disgusting that Sony have waiting 7 days to tell users that their Credit Card details may have been compromisedโ€. โ€ข โ€œI bet the hacker will get emails out quicker than Sony!โ€ 05/08/2016 ยฉ The University of Sheffield 16 Sony PlayStation Network
  • 17. Toyota โ€ข While there was still anger and negative viewpoints shared through social media services, โ€ข Company was able to minimise their impact by eliminating confusion and keeping the consumer base regularly informed of developments 05/08/2016 ยฉ The University of Sheffield 17
  • 18. Brand Management โ€ข The two cases have highlighted brands need to know how they are being mentioned across social media profiles โ€ข Social Media Analytics is now a huge market 05/08/2016 ยฉ The University of Sheffield 18
  • 19. 05/08/2016 ยฉ The University of Sheffield 19 API โ€ข An API in simple terms is a set of requirements which dictate how one application may talk to another โ€ข On the web APIs allow services such as Google maps or Facebook let other application โ€˜piggybackโ€™ on their services โ€ข APIs allow outside programs to access specific sets of features thus saving developers time and resources i.e., accessing data See http://readwrite.com/2013/09/19/api-defined/
  • 20. 05/08/2016 ยฉ The University of Sheffield 20 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข i) Twitterโ€™s Search API โ€“ similar to the search function in Twitter. It is focused on relevance and not completeness, some tweets and users may be missing from results (possible to go back in time 7 days from the query date up to 3200 queries)
  • 21. 05/08/2016 ยฉ The University of Sheffield 21 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข ii) Twitterโ€™s Streaming API โ€“ Streaming APIs give developers low latency access to Twitterโ€™s global stream of tweet data (a type of โ€˜liveโ€™ stream)
  • 22. 05/08/2016 ยฉ The University of Sheffield 22 Different types of API โ€ข Application Programming Interface: Three types of Twitter API i) Search, ii) Stream, and iii) Firehose โ€ข iii) Firehose API โ€“ in theory, 100% of Twitter data (most software allows up to 30 days worth of historical tweets)
  • 23. 05/08/2016 ยฉ The University of Sheffield 23 Legality โ€ข Sharing Twitter datasets is prohibited under Twitterโ€™s terms and conditions see: https://dev.twitter.com/overview/terms/agreeme nt-and-policy โ€ข However, sharing Tweet IDs (to look up the tweets that are used is permissible).
  • 24. 05/08/2016 ยฉ The University of Sheffield 24 Examples of Methods Used to Analyse Twitter Data โ€ข Sentiment analysis has the potential to work well with Twitter data, as tweets are short (i.e., <=140 characters) โ€ข However, sarcasm is difficult to detect within tweets. โ€ข See the SentiStrength algorithm (http://sentistrength.wlv.ac.uk/)
  • 25. 05/08/2016 ยฉ The University of Sheffield 25 Types of Analysis โ€ข Time series analysis can be used to examine tweets overtime and to see when a peak in tweets may occur or have occurred.
  • 26. 05/08/2016 ยฉ The University of Sheffield 26 See Example of DWP
  • 27. Types of Analysis โ€ข Network analysis is used to visualize the connections between users and visually display the connections โ€ข Network analysis can also be used to cluster Twitter users into different groups 05/08/2016 ยฉ The University of Sheffield 27
  • 28. Types of Analysis โ€ข Machine Learning โ€ข Example: Manually coding a subset of data e.g., 100 tweets in a dataset of a 1,000 where the algorithm will automatically classify the remaining data. 05/08/2016 ยฉ The University of Sheffield 28
  • 29. Types of Analysis โ€ข Content Analysis/Thematic Analysis โ€ข Example: Manually placing tweets into predefined categories and/or exploring tweets and developing categories. โ€ข Use Microsoft Excel/NVivo 05/08/2016 ยฉ The University of Sheffield 29
  • 30. Part 2 โ€ข Part 2 of the seminar will provide an overview of some of the most popular social media analysis and analytical software available to the public. 05/08/2016 ยฉ The University of Sheffield 30
  • 31. Visibrain 05/08/2016 ยฉ The University of Sheffield 31
  • 32. Visibrain โ€ข Visibrain is a powerful media monitoring tool which has access to the Twitter Firehose โ€ข Therefore, it is not limited in the amount that can be retrieved unlike the Search and Stream APIs. 05/08/2016 ยฉ The University of Sheffield 32
  • 33. Visibrain โ€ข Explore and filter tweets, links, users, hashtags, expressions, mentions, countries, languages, genders, audiences, occupations, interests, client types, content types, and domains. 05/08/2016 ยฉ The University of Sheffield 33
  • 34. Possible uses โ€ข Security teams, organisations, and/or government โ€ข For example you could monitor Twitter/social media for mention of a bank and a hacking group and/or a city and โ€˜attackโ€™ โ€ข If triggered you would receive an alert 05/08/2016 ยฉ The University of Sheffield 34
  • 35. Visibrain 05/08/2016 ยฉ The University of Sheffield 35
  • 36. Visibrain โ€ข Receive reports produced and delivered almost instantly displaying tweet content, actors, URLS, and so forth. โ€ข Possible to receive these alerts via email to almost any location in the world. 05/08/2016 ยฉ The University of Sheffield 36
  • 37. Visibrain 05/08/2016 ยฉ The University of Sheffield 37
  • 38. Echosec 05/08/2016 ยฉ The University of Sheffield 38
  • 39. Echosec โ€ข Many citizens now carry mobile devices connected to the Internet capable of taking video and pictures โ€ข Emerging news stories of crisis events may be reported on social media before reported anywhere else. These posts may contain location information. 05/08/2016 ยฉ The University of Sheffield 39
  • 40. Echosec โ€ข For example, those whom are affected by a flood may post information and media with location information such as video or pictures to social media which can then be utilized by the emergency services 05/08/2016 ยฉ The University of Sheffield 40
  • 41. Echosec โ€ข Echosec has removed the difficulty of identifying relevant content by providing the ability to search for locations as well as content 05/08/2016 ยฉ The University of Sheffield 41
  • 42. Echosec โ€ข Echosec allows you to navigate to almost any location in the world and examine the social media activity around that area โ€ข Currently Echosec Pro allows users to access at least the following social data feeds: Instagram, Twitter, Foursquare, AIS Shipping, Sina Weibo, Flickr, YouTube, VK, and Panoramio 05/08/2016 ยฉ The University of Sheffield 42
  • 43. Possible uses of Echosec โ€ข Crisis events: โ€ข Floods โ€ข Earthquakes โ€ข Situational awareness: โ€ข Concerts โ€ข Sports events โ€ข Security Teams: โ€ข Government security teams/ Local Authorities โ€ข Private Investigators 05/08/2016 ยฉ The University of Sheffield 43
  • 44. Possible uses of Echosec 05/08/2016 ยฉ The University of Sheffield 44 โ€ข Geographically searching social media data in an area can show you what you are looking for in an emergency โ€ข Can examine locations of affected areas and see where people have stopped posting from
  • 45. Possible Use Case 05/08/2016 ยฉ The University of Sheffield 45 โ€ข Security services e.g., NSA examining content shared from building premises for security breaches โ€ข Echosec has found cases of employees working within secure buildings to be taking selfies with their ID badges
  • 46. Dashboard layout of Echosec 05/08/2016 ยฉ The University of Sheffield 46
  • 47. NodeXL 05/08/2016 ยฉ The University of Sheffield 47
  • 48. NodeXL โ€ข Social media analysis that looks at the structure of networks โ€ข Produces analytics such as most frequently occurring URLs, domains, hashtags, keywords, overall and by group level. 05/08/2016 ยฉ The University of Sheffield 48
  • 49. NodeXL โ€ข Social media analysis that looks at the structure of networks โ€ข Produces analytics such as most frequently occurring URLs, domains, hashtags, keywords, overall and by group level. 05/08/2016 ยฉ The University of Sheffield 49
  • 50. NodeXL โ€ข Essentially NodeXL allows you to take pictures of crowds โ€ข Graph Gallery โ€“ a โ€œFlickr for network graphsโ€ โ€“ Marc Smith (Social Media research foundation) โ€ข DWP network graph by Connected Action Consulting Group 05/08/2016 ยฉ The University of Sheffield 50
  • 51. NodeXL 05/08/2016 ยฉ The University of Sheffield 51
  • 52. NodeXL โ€“ 6 types of network structures 05/08/2016 ยฉ The University of Sheffield 52
  • 53. NodeXL โ€“ 6 types of network structures 05/08/2016 ยฉ The University of Sheffield 53
  • 54. 05/08/2016 ยฉ The University of Sheffield 54 NodeXL
  • 55. 05/08/2016 ยฉ The University of Sheffield 55 NodeXL interactive graph
  • 56. 05/08/2016 ยฉ The University of Sheffield 56 NodeXL G1 MP for Ashford. Secretary of State for Work and Pensions.
  • 57. 05/08/2016 ยฉ The University of Sheffield 57 NodeXL G2 DWP Twitter account
  • 58. NodeXL โ€ข [640] Words in Sentiment List#1: Positive โ€ข [1544] Words in Sentiment List#2: Negative โ€ข [4] Words in Sentiment List#3: Angry/Violent 05/08/2016 ยฉ The University of Sheffield 58
  • 59. NodeXL โ€ข Top Word Pairs in Tweet in Entire Graph: [325] phone,lines [319] lines,dwp [214] scrap,rip [196] rip,phone [169] expensive,phone [169] ministry,labour [159] calls,dwp [139] scrap,dwp [137] dwp,minister [128] new,dwp 05/08/2016 ยฉ The University of Sheffield 59
  • 60. NodeXL โ€ข [160] Social Security Advisory Committee, Telephony in DWP and HMRC: an update https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/5 38834/telephony-in-dwp-and-hmrc-an-update-op17.pdf โ€ข [68] Benefit sanctions handed out to thousands of people โ€˜unlawfulโ€™ http://www.independent.co.uk/news/uk/politics/dwp-defeated-in-court-again- over-unlawful-retroactive-benefit-sanctions-a7006466.html โ€ข [32] Make all calls to the department for work and pensions free https://you.38degrees.org.uk/petitions/make-ian-duncan-smith-make-all-phone- communication-free?bucket=blast&source=twitter-share-button โ€ข [27] Owen Smith vows to scrap the DWP as โ€˜cold-eyed revolutionโ€™ in Labour leadership speech http://www.mirror.co.uk/news/uk-news/owen-smith- labour-leadership-revolution-8502537 05/08/2016 ยฉ The University of Sheffield 60
  • 61. NodeXL 05/08/2016 ยฉ The University of Sheffield 61 โ€ข Although the graph gallery contains a number of network graphs from Twitter - it is possible to import networks from other social media platforms such as Facebook and Instagram
  • 62. DiscoverText 05/08/2016 ยฉ The University of Sheffield 62
  • 63. 05/08/2016 ยฉ The University of Sheffield 63 DiscoverText โ€ข Useful for machine learning, de- duplication, and collaboration. โ€ข Used with Sifter, it is possible to retrieve data going back in time more than 30 days โ€ข Data cost will range from not that expensive to very expensive depending on the query and time
  • 64. 05/08/2016 ยฉ The University of Sheffield 64 DiscoverText โ€ข University of Sheffield Project โ€ข Analysed the mentions of the University over a 1 year period and classified tweets โ€ข Better understand who engages with the University on Twitter
  • 65. 05/08/2016 ยฉ The University of Sheffield 65 Conclusion โ€ข There is no best social media analytics or analysis software โ€“ each offers something different. โ€ข I tend to use a number of software in combination to gain the best insight. โ€ข Thanks for listening.
  • 66. 05/08/2016 ยฉ The University of Sheffield 66 Questions?