harnessing unstructured data with text mining
play

Harnessing Unstructured Data with Text Mining Jarlath Quinn - PowerPoint PPT Presentation

Harnessing Unstructured Data with Text Mining Jarlath Quinn Analytics Consultant Rachel Clinton Business Development www.sv-europe.com A SELECT INTERNATIONAL COMPANY FAQs Is this session being recorded? No Can I get a copy


  1. Harnessing Unstructured Data with Text Mining Jarlath Quinn – Analytics Consultant Rachel Clinton – Business Development www.sv-europe.com A SELECT INTERNATIONAL COMPANY

  2. FAQ’s • Is this session being recorded? No • Can I get a copy of the slides? Yes, we’ll email a PDF copy to you after the session has ended. • Can we arrange a re-run for colleagues? Yes, just ask us. • How can I ask questions? All lines are muted so please use the chat facility – if we run out of time we will follow up with you. A SELECT INTERNATIONAL COMPANY

  3. Agenda • About Smart Vision Europe • Overview of Text Mining • Definitions of data availability and types • Analysing news and media output • Understanding customer sentiment • Categorising customer comment fields • Enhancing traditional analytics with unstructured information A SELECT INTERNATIONAL COMPANY

  4. • Premium, accredited partner to IBM specialising in the SPSS Advanced Analytics suite. • Team each has 15 to 20 years of experience working in the predictive analytic space - specifically as senior members of the heritage SPSS team A SELECT INTERNATIONAL COMPANY

  5. Unstructured Data • 90% of the world’s data was generated in the last two years and 80% of that data is unstructured - IBM A SELECT INTERNATIONAL COMPANY

  6. Trending Applications • Google trends 2007 - 2013 Sentiment Analysis Social Media Analytics A SELECT INTERNATIONAL COMPANY

  7. Driving value from data… Profit / Effectiveness Social Media Interaction Data Descriptive Data Transactional data Insight A SELECT INTERNATIONAL COMPANY

  8. Utilising a powerful, proven methodology • CRISP-DM: Cross-Industry Standard Process for Data Mining • Each application can be developed and progressed through a series of key phases A SELECT INTERNATIONAL COMPANY

  9. Using Text Data to Address Key Objectives attract grow risk retain A SELECT INTERNATIONAL COMPANY

  10. Using Text Data to Address Key Concerns • Social Media • Brand perception attract • Company reputation • Reaction to advertisements/campaigns • Call centre notes, emails • Enquiries from potential customers • Cancellation statements • Web • Website search terms A SELECT INTERNATIONAL COMPANY

  11. Using Text Data to Address Key Concerns • Call centre notes, emails • Enquiries from existing customers grow • Enhancing existing models • Web • Website search terms • FAQ’s A SELECT INTERNATIONAL COMPANY

  12. Using Text Data to Address Key Concerns • Call centre notes, emails • Enquiries from existing customers retain • Complaints • Web • Website searches • Surveys • Feedback • Cancellation emails • Social Media • Negative sentiment A SELECT INTERNATIONAL COMPANY

  13. Using Text Data to Address Key Concerns • Assets • Computer Logs/ Browsing history risk • Web • Website searches • Call centre notes, emails • Insurance claims, fault notifications • Collusion • Social Media • Discussion of scams, fraud • RSS Feeds • Competitor/market monitoring A SELECT INTERNATIONAL COMPANY

  14. Example Text Analytics Applications • Insight/Sentiment Categorisation • What are the people talking about? • How do they feel about key topics? • Classification • How can we automatically categorise documents in a repository? • Model Enhancement • How can we incorporate text data to make better predictions? • Monitoring • How can we keep up to date with what customers/competitors are saying about us/our products? A SELECT INTERNATIONAL COMPANY

  15. Lets take a deeper look www.sv-europe.com A SELECT INTERNATIONAL COMPANY

  16. Advice to get started • Consider adopting a proven methodology e.g. CRISP-DM (www.CRISP-DM.eu) • What are you trying to achieve? Focus on Business Understanding and Deployment • Develop your own libraries of terms and categories • Consider integration with other business insight systems (e.g. MI/BI) • How will you know its worked? Focus on measuring the benefit – e.g. faster categorisation, more accurate models, more timely decisions • You may not need to recruit specialists: data-literate, business-focussed people can learn how to do this. A SELECT INTERNATIONAL COMPANY

  17. Contact us: +44 (0)207 786 3568 info@sv-europe.com Thank you www.sv-europe.com A SELECT INTERNATIONAL COMPANY

Recommend


More recommend