Social analytics Alok Kumar Jha IBM 2011_may27


Published on

Information Excellence Knowledge Share Session 101 on "Social Media Analytics for creating effective relationship, build advocacy and improve loyalty"

Published in: Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Social analytics Alok Kumar Jha IBM 2011_may27

  1. 1. Social AnalyticsCreating relationship. Build advocacy. Improve loyaltyAlok Kumar JhaBI Solution ArchitectIndia Software Lab, SWGBangalore More than Listening … Social Media Strategy © 2011 IBM Corporation
  2. 2. Social Media Analytics is a specific application of text analytics Slide 2 © 2011 IBM Corporation
  3. 3. Social Media Influence 78% of consumers trust peer recommendations Source: August 25, 2009, “The Broad Reach Of Social Technologies” Forrester report 770 million people worldwide visited a social networking site. Source: com Score, Social Networking Phenomenon Slide 3 © 2011 IBM Corporation
  4. 4. Key Questions…How can we make more money, more cost effectively and quicklyto drive social marketing ROIHow do we engage our customers and stakeholders to enhanceour reputation and brandHow do we transform marketing to create more intimate and lastingrelationships with our customers via social media and influenceadvocacy of our brand Slide 4 © 2011 IBM Corporation
  5. 5. Why Social Media Analytics? Analytics that listen, measure and analyze social media performance to more effectively: Understand your customer needs to target new Grow Your offers and products more cost-effectively through Business different social media channels Evaluate your corporate reputation and make Enhance Your evidence-based messaging decisions that target Reputation the right stakeholders at the right time Respond more quickly with accurate, timely and Improve your relevant insight into customer requests to customer experience ensure a consistent brand experience across all channels Slide 5 © 2011 IBM Corporation
  6. 6. IBM Cognos Consumer Insight Enables marketing professionals to be more precise, agile and responsive to customer demands and opinions expressed through social media by dynamically listening, measuring and analyzing large volumes of publicly available content on the Internet – the only solution to leverage a broader business analytics environment by seamlessly connecting sophisticated packaged reporting and analyses to Cognos BI. Slide 6 © 2011 IBM Corporation
  7. 7. How Does IBM Cognos Consumer Insight Work? •Automatically Process content •Communicate insight identify and tag based on relevance broadly across the relevant content to business business •Admin User Interface Interactive 3rd Party Providers Visualization & •Crawl Analysis •Semantic & Search-led Faceted Data Search Index Multiple Languages •Information Exploration Extraction Cognos Web content •Compute Reports & Infrastructure •Database Analytics Analytics Platform Integrated UI Customers use of content is subject to the terms and conditions of the respective third party data providers Slide 7 © 2011 IBM Corporation
  8. 8. Cognos Consumer Insight Layers Presentation Layer Data Layer Sentiments Concepts Snippets Faceted Search Indexes Hotwords Star Schema Tables Analytics Layer WorkFlow Definition Slide 8 © 2011 IBM Corporation
  9. 9. Cognos Customer Insights (CCI) High Level Platform Architecture Presentation Tier Built-in Capabilities: Additional Capabilities: Analysis UI, Administration UI Cognos BI Reporting, IBM Mashup Center Analysis Tier Crawler Flow Manager IBM System T Search Uploader IBM Hadoop MapReduce / IBM GPFS File System Data Tier Sentiments Brands Analysis DAS Data Loader Snippets Services Hotwords Directory Content Store Star Schema Tables Search Indexes Slide 9 © 2011 IBM Corporation
  10. 10. Presentation Layer Presentation Layer consists of components that provide: • Reporting and Analysis • Exploration, Search and Discovery • CCI AdministrationSlide 10 © 2011 IBM Corporation
  11. 11. CCI Architecture: Presentation TierPresentation Tier Cognos 10 Business Intelligence WAS Community Edition Cognos 10 Mashup Services IBM Mashup Center 3.0 Analysis WebApp Administratior WebApp SLES Linux x86 64-bit 11.0• Exploratory UI to discover consumer sentiment topics and trends• Administration UI to model, configure and schedule analysis jobs• Advanced presentation layer with mashups to combine information from multiple sources• Customizable set of IBM Cognos BI reports and dashboards with drill up / down capability Slide 11 © 2011 IBM Corporation
  12. 12. Cognos Consumer Insight Layers Presentation Layer Data Layer Sentiments Concepts Snippets Faceted Search Indexes Hotwords Use Presentation layer (Admin UI) Star Schema Tables to submit analytics workflows to Analytics Layer Analytics layer. WorkFlow Definition Slide 12 © 2011 IBM Corporation
  13. 13. Analytics Layer WorkFlow Definition Analytics Layer consists of components that provide: • Map/Reduce framework to manage multiple concurrent jobs • Text Analytics, Entity Extraction and Sentiment Analysis • Topics Discovery and Detection • Extraction and Transformations Slide 13 © 2011 IBM Corporation
  14. 14. CCI Architecture: Analysis Tier Analysis Tier WAS Community Edition SPSS Modeler 19 SPSS Text Analytics Workflow Definitions System T IBM Hadoop / MapReduce Crawler Flow Manager JAQL Uploader LanguageWare SLES Linux x86 64-bit 11.0 • Distributed processing for information analysis using IBM Hadoop MapReduce framework • Crawlers extract data from the web and load it into Hadoop • Mapper tasks run in parallel each processing portions of overall data • Reduce tasks aggregate mapper task data and feed into another chain of mapper or reduce tasks • System T applies Brand, Keyword, Sentiment and Named Entities detection rules • JAQL used as scripting language to abstract Hadoop processing Slide 14 © 2011 IBM Corporation
  15. 15. Cognos Consumer Insight Layers Presentation Layer Data Layer Sentiments Concepts Analytics layer computes results and Snippets Faceted Search supplies data to the Data layer. Indexes Hotwords Star Schema Tables Analytics Layer WorkFlow Definition Slide 15 © 2011 IBM Corporation
  16. 16. Data Layer Sentiments Brands Snippets Faceted Search Indexes Hotwords Star Schema Tables Data Layer consists of components that provide: • ETL of analysis results from Analytics layer • Star schema fact and dimension tables • Facetted search indexes and correlation computations • Report Model for drill-up, drill-down, drill-thru definitionsSlide 16 © 2011 IBM Corporation
  17. 17. Cognos Consumer Insight Layers Presentation Layer Data Layer Sentiments Concepts Snippets Faceted Search Indexes Hotwords Star Schema Tables Analytics Layer Data layer provides formatted data for WorkFlow Definition Presentation layer to render into UIs. Slide 17 © 2011 IBM Corporation
  18. 18. CCI Architecture: Data Tier Data Tier DB2 Workgroup Server 9.7 WAS Community Edition Sentiments Brands Analysis DAS Snippets Data Access ILEL Hotwords Services WebApp IBM Lucene Indexing Admin Schema Staging Area Star Schema Data Loader Directory Services OpenLDAP 2-2.4.12- Cognos Content Store 7.19.1 SLES Linux x86 64-bit 11.0 • Services to store, index and access information collected during analysis • Services to compute statistics and affinity measures • Configuration repository for CCI workspaces • Analysis DAS retrieves and caches faceted search results and aggregations from Lucene indexes • Analysis DAS builds topic vs. date JSON records used by Analysis UI to render topic rivers • Cognos BI Server and SPSS consume data from the star schema tables Slide 18 © 2011 IBM Corporation
  19. 19. Putting it all together – End-to-End Process 1 Collect 2 Analyze 3 Deliver Crawler JAQL - System T Uploader boards • Sentiment Lucene BoardReader Analysis Index blogs • Entity Detection • Topic Detection news DB2 … … •Adhoc jobs usuallly used during implementation •Scheduled jobs (daily, weekly, monthly) used in once queries are defined Slide 19 © 2011 IBM Corporation
  20. 20. Questions? ? ? ? Slide 20 © 2011 IBM Corporation
  21. 21. © Copyright IBM Corporation 2008 All rights reserved. The information contained in these materials is provided for informational purposes only, and is provided AS IS w ithout w arranty of anykind, express or implied. IBM shall not be responsible for any damages arising out of the use of, or otherw ise related to, these materials. Nothing contained in these materials is intended to, norshall have the effect of, creating any w arranties or representations from IBM or its suppliers or licensors, or altering the terms and conditions of the applicable license agreement governing the useof IBM softw are. References in these materials to IBM products, programs, or services do not imply that they w ill be available in all countries in w hich IBM operates. Product release dates and/orcapabilities referenced in these materials may change at any time at IBM’s sole discretion based on market opportunities or other factors, and are not intended to be a commitment to future productor feature availability in any w ay. IBM, the IBM logo, Cognos, the Cognos logo, and other IBM products and services are trademarks of the International Business Machines Corporation, in theUnited States, other countries or both. Other company, product, or service names may be trademarks or service marks of others. Slide 21 © 2011 IBM Corporation
  22. 22. IBM Cognos BI Reports • Use IBM Cognos 8 BI reports to perform more advanced reporting and obtain a summary view of your data • CCI includes four pre- defined reports found in tabbed portal pages: Coverage, Hotword, Analysis, Influencers and Sentiment • Drill from the report to exploratory analysis portal drill to the text of the snippets Slide 22 © 2011 IBM Corporation
  23. 23. Coverage Slide 23 © 2011 IBM Corporation
  24. 24. Hotword Slide 24 © 2011 IBM Corporation
  25. 25. Influencers Slide 25 © 2011 IBM Corporation
  26. 26. Sentiment Slide 26 © 2011 IBM Corporation