-Enrichment - Unlocking the value of data for digital transformation - Big Data Expo 2019
1 like178 views
The document discusses the significance of big data in driving digital transformation, emphasizing its rise due to the increasing volume, variety, and value it presents for businesses. It outlines challenges like data quality and security while proposing a structured approach to develop smart data strategies through identification, storage, provision, processing, and governance of data. Ultimately, smart data, characterized by aggregation, normalization, and enrichment, is presented as essential for maximizing business insights and profitability.
-Enrichment - Unlocking the value of data for digital transformation - Big Data Expo 2019
- 1. Unlocking Big Data to Power Digital Transformation How data enrichment gets you to smart data.
- 2. 2 Today’s Agenda 1. Why is big data on the rise? 2. How is big data connected to the business mandate of digital transformation? 3. What big data challenges need to be overcome to achieve measurable value 4. Where does a smart data process fit into digital transformation?
- 3. 3 Why is big data on the rise? Data itself isn’t new. The Ishango Bone - Tally Stick from 18,0000 BC Egyptian Hieroglyphics - Early data visualizations
- 4. 4 Why is big data on the rise? Data empowers insight • Meets consumers’ data needs • Monetizes data for ad relevance • Data exhaust reveals important information Google’s ‘Aha’ moment offers a clue.
- 5. 5 What is data exhaust? The breadcrumbs of internet searches speak volumes. Every click counts • Patterns of search terms • Query phrasing • Spelling & punctuation • Location • Dwell times • Click patterns
- 6. 6 How does Google use data exhaust? Improving search relevancy with location data is just one benefit. Find Stroopwafels near me
- 7. 7 Growing value of data Data is intrinsic to digital transformation. Digital transformation creates more data. $1 Trillion invested in digital transformation in 2018 Poor data quality costs businesses Millions every year
- 8. 8 Data leaders unlock data’s value Those who don’t will be left behind. Google may have led the way, but these days: 80%of the most innovative companies use data to drive advantages in their business
- 9. 9 The Internet of Things (IoT) also multiplies the data available Making use of IoT data has powerful potential. Connected sneakers, wearables and apps help Nike understand runners’ behaviors.
- 10. 10 The Internet of Things (IoT) also multiplies the data available Making use of IoT data has powerful potential. Propeller Health’s use of user sensor data led to a 50% reduction in asthma attacks.
- 11. 11 The Internet of Things (IoT) also multiplies the data available Making use of IoT data has powerful potential. On-car sensors help teams adjust race strategies in real time to bring home the win.
- 12. 12LexisNexis Confidential “Companies are either: • Already data companies • Will become data companies • Or will cease to exist.” ~ Stephen Brobst, CTO of Teradata Are we there yet? Driving toward digital transformation is a must.
- 13. 13LexisNexis Confidential 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Not Important Somewhat Important Important Very Imortant Critical Examples of Digital Transformation Where is digital transformation already well-established?
- 14. 14LexisNexis Confidential Digital innovation and data are transforming industries Consumers want personalized, go-everywhere experiences.
- 15. 15LexisNexis Confidential Digital innovation and data are transforming industries Digitization is a powerful ally of the empowered employee.
- 16. 16LexisNexis Confidential Digital innovation and data are transforming industries No industry is immune.
- 17. Of knowledge workers' work-days are spent trying to find and manage data 17 2.5 Quintillion Bytes of data is created every day 5 Billion Inquiries are processed by search engines daily 90% Of All Data has been generated in the past 2 years Source: http://bit.ly/3154f13 Of companies are dissatisfied with their ability to filter irrelevant data 60% 43% Source: http://bit.ly/2LPV01t The Challenges: Volume The volume of data has increased exponentially.
- 18. 18 The Challenges: Variety The variety of data types and sources has also exploded. Can you use more data types? • News • Audio & video • Social
- 19. 19LexisNexis Confidential The Challenges: Veracity Validating the trustworthiness of data poses problems. Can you trust the data? • Media bias • Typos • Colloquialisms
- 20. 20 The Challenges: Velocity Keeping up with the constant data barrage. Can you take advantage of real-time data? • Clicks and calls • 24/7/365 news
- 21. 21LexisNexis Confidential The Challenges: Vulnerability With great power comes great responsibility. Can you protect your data? • Privacy concerns of consumers • Security expectations of regulators
- 22. 22 5 key components for a data strategy Identify. Store. Provision. Process. Govern GovernProcessProvisionStoreIdentify
- 23. 23 Component 1: Identify Identify data and understand its meaning regardless of structure, origin or location. Search for the right data
- 24. 24 Component 2: Store Ensure the data you gather is easy to access and process. Determine how to store the growing volume of data
- 25. 25 Component 3: Provision Set rules and access guidelines for data. Pack up the data so it can be reused and shared
- 26. 26 Component 4: Process Establish how different types of data will be processed for maximum value. Gather data and process it for a unified, consistent view
- 27. 27 Component 5: Govern Create policies for using and protecting data. Use data responsibly
- 28. 28 Moving from big data to smart data Get maximum Value from data to support digital transformation. better results. Smarter data helps deliver
- 29. 29 Revealing the true value of big data What makes data smart? • Aggregation • Normalization • Enrichment
- 30. 30 INTERNAL Datasets CURATED Datasets PUBLICLY AVAILABLE Datasets Integrated with Data Analysis Tools Data aggregation: Layered data yields better insights Build variety by going beyond the confines of internal datasets.
- 31. 31 Data normalization: Creating clean, semi-structured data Make data more searchable and user-friendly. Structured versus Unstructured
- 32. 32 Data enrichment: Enhancing the quality of data with tags Metadata describes other data, making it easier to search. Structured versus Unstructured
- 33. 33 Our Smart Data Process Licensing of news & social commentary, company & industry information, legal & patents data, people data & public records Aggregation, normalization & archival of data using cloud technologies Enrichments to enhance data value while reducing data volume Cloud-based applications and Data as a Service options to meet a wide range of use cases 2 1 3 4
- 34. 34 Entity Extraction & Resolution Improves ability to narrow a search to a specific entity.
- 35. 35 Classification Enables search refinements to narrow results to the most relevant data
- 36. 36 Sentiment Understand and anticipate trends, analyze value of media coverage and more.
- 37. 37 Identifier Metadata Further refine data searches to reduce volume while increasing relevance
- 38. 38 Smart data puts digital transformation on a fast track “Businesses that successfully apply artificial intelligence (AI) could increase profitability by an average of 38 percent by 2035.” — Accenture Research Descriptive Analytics What has happened? • Using data aggregation and data curation to provide insight into the past • Analyzing historical data to report on financials, inventory, customers
- 39. 39 Smart data puts digital transformation on a fast track “Businesses that successfully apply artificial intelligence (AI) could increase profitability by an average of 38 percent by 2035.” — Accenture Research Predictive Analytics What could happen? • Using statistical models and forecasts techniques to understand the future • Analyzing purchases and sentiment to predict the next big consumer trend
- 40. 40 Smart data puts digital transformation on a fast track “Businesses that successfully apply artificial intelligence (AI) could increase profitability by an average of 38 percent by 2035.” — Accenture Research Prescriptive Analytics What should we do? • Using optimization and simulation algorithms to advise on possible outcomes • Automated stock trades based on quant models
- 41. Here’s to a Smart Data strategy & one last ‘V’: Victory! Thank You