Social Media Mining and Analytics
Download as PPTX, PDF3 likes1,753 views
The document explores social media from a scientific perspective, highlighting its definition, various types, and the challenges of data overload. It discusses applications such as graph mining, sentiment analysis, and social media engagement in fields like marketing and journalism. Additionally, it touches on the impact of social media on traditional media and its relevance in areas such as military operations and election campaigns.
![Introduction“In contrast to one-to-many communication structure of traditional mass-media, social media allows the emergence of many-to-many communication, and gives rise to a mass of self-communication” [Castells, 2009]6https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011](https://image.slidesharecdn.com/socm-110920232003-phpapp02/85/Social-Media-Mining-and-Analytics-6-320.jpg)
![Power Law: Research On Twitter CharacteristicsThe number of followings (solid line) and that of followers (dotted line) on Twitter [Kwak et al., 2010]21https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011](https://image.slidesharecdn.com/socm-110920232003-phpapp02/85/Social-Media-Mining-and-Analytics-21-320.jpg)
![Social Media Effects on Traditional Media [An et al., 2011]Social media changes the ways in which peopleRead newsShare newsInteract on news items“News no longer breaks, it tweets”22https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011](https://image.slidesharecdn.com/socm-110920232003-phpapp02/85/Social-Media-Mining-and-Analytics-22-320.jpg)
![New DirectionsFirst International Workshop on Social Media Engagement (SoME011, in conjunction with WWW2011)Engagement defines the phenomena of being captivated and motivated.Engagement can be measured in terms of a single interactive session or of a more long-term relationship with the social platform across multiple interactions.To design not just systems, but rather engaging experiences.[Attfield et al., 2011] “Towards a science of user engagement ”.34https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011](https://image.slidesharecdn.com/socm-110920232003-phpapp02/85/Social-Media-Mining-and-Analytics-34-320.jpg)
Social Media Mining and Analytics
- 1. Social MEDIA MINING AND ANALYTICSA Look into Social Media from a Scientific Perspective1https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 2. IntroductionWhat is Social Media?2https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 3. IntroductionSocial media is the hot thing in town. Everyone wants to do it. No one actually knows how. (AvinashKaushik, Google’s analytics evangelist)3https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 4. IntroductionSocial Media is not only about Social Networks4https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 5. IntroductionSocial Media is a media for social interaction using highly accessible and scalable communication techniques.Social Media is the use of web-based and mobile technologies to turn communications into interactive dialog.- Wikipedia5https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 6. Introduction“In contrast to one-to-many communication structure of traditional mass-media, social media allows the emergence of many-to-many communication, and gives rise to a mass of self-communication” [Castells, 2009]6https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 7. Social Media ExamplesSocial communication: emails, mobiles, forums, chatsSocial networking: facebook, google+Social blogging/microblogging: twitter, livejournal, bloggerSocial sharing: flickr, vimeo, youtubeSocial news: digg, slashdot, cnn, ireportSocial bookmarking: delicious, citulikeSocial knowledge/wikis: wikipedia, tripadvisorSocial shopping: groupon, amazon, ebaySocial apps and games: foursquare, farmville7https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 8. Social Media ChallengeToo much data!8https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 9. Social Media GoalsWhat could we do?Ask the right questions9https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 10. QuestionsWhy social media?What is social media for?Social Media Policies and StrategiesWhat is going on?What can we learn?How to exploit social media?Social Media MiningSocial Media Practices10https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 11. Talk OrganizationGraph MiningOpinion Mining and Sentiment AnalysisSmart MobilitySocial Media Engagement and Social Innovations11https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 12. Graph Mining12https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 13. Graph Mining: IntroductionSocial Media social presence, social interactionsGraph G=G(V,E)V is the set of vertices, or nodesE is the set of edges which may or may not have weights13https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 14. Example‘user user’ graphs on the base of social interactions (e.g. friendship, communications: sharing, commenting)14https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 15. Example‘user properties’ bipartitegraphs15https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 16. Quiz16https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 17. Quiz17https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 18. What is the difference?18https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 19. Power LawPower law is a special family of distributionsHuman heightsCity populationNo. of books sold19https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 20. Power LawRandom variable X has a power law distribution with exponent α: P (X>x) ~ x-αas x∞Pareto principle: for many events roughly 80% of the effects come from 20% of the causes20https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 21. Power Law: Research On Twitter CharacteristicsThe number of followings (solid line) and that of followers (dotted line) on Twitter [Kwak et al., 2010]21https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 22. Social Media Effects on Traditional Media [An et al., 2011]Social media changes the ways in which peopleRead newsShare newsInteract on news items“News no longer breaks, it tweets”22https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 23. Media Graph Mining Results: Computational Journalism Journalists are playing prominent role in social mediaAudience interactionPersonal update11K23https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 24. Media Graph Mining Results: Media Publishers Reach More AudienceYes: Social interaction increases publisher’s audienceOn average, audience size increases by a factor of 2824https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 25. Media Graph Mining Results: Users Exposed to Diverse OpinionsYes: Users are exposed to diverse opinions throughsocial interactions25https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 26. Opinion Mining and Sentiment Analysis26https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 27. Terminology and ProblemsSentiment analysis aka opinion mining is the task of automatically detecting sentiment in textActive research area since 2002Standard part of online market research toolkitsOpinions are personal judgements about somethingIt is good. It is bad. It is expensive.Subjective text contains opinions; Objective text states only facts.Sentiments are expressions of emotion or attitude or opinionIt is good. It is bad. I like it. I am happy. I am depressed. I am angry at you.Sentiment analysis is often thought of as the prediction of people’s private/internal states from text27https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 28. Commercial Sentiment Analysis GoalsDetermine overall opinions about a productE.g, the M90 phone is excellentE.g., the M90 phone is expensive but excellentDetermine opinions about parts of a productE.g., the screen of the M90 is too small but its weight is very lightI love the steering wheel on the new Picasso28https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 29. Commercial Sentiment Analysis GoalsDetermine changes in overall customer brand opinion (e.g. daily proportions of positive/negative comments)In response to advertisingAs routine monitoringIdentify individual unhappy customersE.g., identify Tweets that mention the brand and are negative29https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 30. Social Science Sentiment Analysis GoalsTrack trends in sentiment over timeIdentify changes in sentimentDiscover patterns in sentiment use in a communicaton mediumE.g, gender, age, nationality differencesDo women/Russians use more sentiment?30https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 31. Smart Mobility31https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 32. Tesco – Home Plus Korea Case32https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 33. Social Media Engagement and Social Innovations33https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 34. New DirectionsFirst International Workshop on Social Media Engagement (SoME011, in conjunction with WWW2011)Engagement defines the phenomena of being captivated and motivated.Engagement can be measured in terms of a single interactive session or of a more long-term relationship with the social platform across multiple interactions.To design not just systems, but rather engaging experiences.[Attfield et al., 2011] “Towards a science of user engagement ”.34https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 35. Engagement and DevelopmentHow to use social media to focus on social problems? And how to make social media available for the people who generally do not have access to the Internet?35https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 36. Social Networks for Agro-Produce MarketingResearch by DhirubhaiAmbani Institute of Information and Communication TechnologyUse of social network platform such as Twitter for agro-produce marketing: A platform for farmers to sell their product(s) and merchants to buy agro-produce and re-sell in the market. Collected tweets from sellers and buyers can be used to generate classified summarized information36https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 37. Applications37https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
- 38. Applications of Social Media Analytics https://www.facebook.com/WebSTIBA Twitter: #csibaseminar201138Startups built around Social Media Analytics: Thriving Business ModelSocialFlowSocial Media Optimization PlatformWorks in Domains of Viral and Word-of-Mouth MarketingProvides Services to Major Media Outlets Recent studyHow different audiences consumed and rebroadcast messages news organizations were sending out: AlJazeera English, BBC News, CNN, The Economist, Fox News and New York Times
- 39. Applications of Social Media Analytics https://www.facebook.com/WebSTIBA Twitter: #csibaseminar201139Startups built around Social Media Analytics: Thriving Business ModelSocialFlowSocial Media Optimization PlatformWorks in Domains of Viral and Word-of-Mouth MarketingProvides Services to Major Media Outlets Recent studyHow different audiences consumed and rebroadcast messages news organizations were sending out: AlJazeera English, BBC News, CNN, The Economist, Fox News and New York Times
- 40. Applications of Social Media Analytics https://www.facebook.com/WebSTIBA Twitter: #csibaseminar201140Social Media within Military and DefenceThe Pentagon sees the advances in Social Media as the new battlefield, and something that needs to be looked at to track the information being provided about military operations on the web.
- 41. Applications of Social Media Analytics https://www.facebook.com/WebSTIBA Twitter: #csibaseminar201141Election Campaigns
- 42. Thank YouReach me on Twitter: @ArjumandYounusEmail me: ayounus@iba.edu.pk42https://www.facebook.com/WebSTIBA Twitter: #csibaseminar2011
Editor's Notes
- #12: All these topics consider social media from a Web Science perspective and take up social media which from a viewpoint which may have not been heard of in Pakistan
- #20: A power law is a special kind of mathematical relationship between two quantities. When the frequency of an event varies as a power of some attribute of that event (e.g. its size), the frequency is said to follow a power law. For instance, the number of cities having a certain population size is found to vary as a power of the size of the population, and hence follows a power law.
- #21: A power law is a special kind of mathematical relationship between two quantities. When the frequency of an event varies as a power of some attribute of that event (e.g. its size), the frequency is said to follow a power law. For instance, the number of cities having a certain population size is found to vary as a power of the size of the population, and hence follows a power law. Alpha between 1 and 2; finite mean and infinite variance
- #24: Data collection Used previously collected dataset of twitter which is near-complete (ICWSM 2009, Cha. et el] Focus on 80 media sources English-based media A total of 14M followers and their connections (1.2B links, 350,000 tweetsFocus on 80 media sources English-based media A total of 14M followers and their connections (1.2B links, 350,000 tweetsOne of the prominent features of social media:Social media journalists
- #37: 1. What is the maximum, minimum, average price of the product expected by sellers in a specific region? 2. What is the maximum, minimum, average price buyers are willing to pay for a particular product in a specific region? 3. Where should sellers sell to get buyers who are ready to pay more than average expectation of sellers? 4. From where buyers should buy to get sellers, who expect less than average expectation of buyers?
- #41: DARPA planned to spend $42 million (£25m) on the Social Media in Strategic Communication (SMISC) program, with prospective contractors asked to test algorithms through "experiments" with social media, it said.
- #42: DARPA planned to spend $42 million (£25m) on the Social Media in Strategic Communication (SMISC) program, with prospective contractors asked to test algorithms through "experiments" with social media, it said.