Oracle Data Minerハンズオンセミナー170927：③Business Analytics概要と事例

Editor's Notes

• #14: 11500人×4%×年間授業料50万円（仮設定）= 2.3億円（２００万ドル）
• #15: Lets now look at a company that has truly leveraged exploration with Oracle: Valdosta State University. Voldosta Implemented Oracle Business Intelligence Enterprise Edition to provide more than 600 faculty and staff members with rich insight into student academic performance They now use BI Cloud service and Provided instant, on-demand access to student class loads and grades to enable faculty to make important timely decisions on how to facilitate student success. They use Interactive dashboards to improve administrative productivity by more than 500% by enabling administrative staff to fulfill 90% of all data requests from faculty, department heads, deans, and others. Oracle has Enabled the university to examine granular data to draw previously unknown correlations to make more informed decisions—for example, determining that students who eat breakfast on campus have a 10% higher retention rate! They Streamlined IT responsibilities and cut database administration time in half, enabling database administrators and programmers to focus on core tasks, such as implementing new web applications, exploring new technologies, and building new Innovative analytic reports Recently purchased BICS {INFO BELOW FROM Http://www.oracle.com/us/corporate/customers/customersearch/vsu-1-endeca-ss-2156270.html}
• #18: 主な事業 ダンキンドーナツ バスキン・ロビンス （別名サーティワンアイスクリーム） 設立 1953年 本社所在地 アメリカ合衆国マサチューセッツ州キャントン
• #20: DX Marketing is a small but comprehensive targeting marketing firms serving regional and national customers out of Savannah, GA.   Challenge: Improve consumer acquisitions and take advantage of the ever-increasing deluge of data produced by consumers while protecting data privacy and security DX Marketing’s streamlined IT department decided to replace the foundation of its marketing and advertising platform for a more economical and secure cloud environment DX Marketing sought to reduce risk management and infrastructure costs; adopt a wider range  of innovative database features Ability to scale capacity fluidly in order to meet escalating customer demand Solution Oracle Database Cloud Service and Oracle Marketing Cloud Service to take advantage of greater performance, security and scalability compared to its previous computing environment and target marketing DX Marketing houses their 260 million and growing licensed records of U.S consumers in Oracle’s Platform as a Service, specifically the Database Cloud Service. The consumer records are basically anyone in the entire United States that has made a purchase/applied for a bank account with 800 variables detailing demographics, incomes, credit scores, addresses, interests, and preferences. Built in the Oracle Database Cloud Service are Advance Analytics to build predictive behavior models and Data Visualization Cloud Services to intuitively analyze customer and sales data opening the opportunity for DX to hire it’s first data scientist to explore the treasure trove of consumer data Advanced cloud tooling capabilities to streamline time-consuming administrative tasks Optimal environment for a small and medium sized business budget DX Marketing expanded their cloud journey with Oracle Data as a Service (DaaS) that is a result of BlueKai and DataLogix acquisitions and is now the largest digital exchange in the world with over 45,000 personal interest variables and attributes. There are B2C and B2B categorizes and industries with optimized anonymous internet browsing data using cross-channel and ID mapping to link fragmented customer touchpoints from multiple devices. This provides a rich, connected profile of intent signals, interests, sentiment, themes, and topics across online, mobile, search, social, and video channels. DX Marketing also the Oracle Software as a Service (SaaS) Marketing Cloud’s BlueKai and Responsys to execute, nurture, and analyze multi-channel campaigns in combination with the massive offline consumer records and predictive behavior modeling. Results In addition to providing DX Marketing with a reliable, secure, and scalable cloud environment  Oracle Cloud Services have enabled the company to decrease time to market for customer campaigns by 70 percent.   Furthermore, the Oracle Cloud solutions have enabled DX Marketing to streamline administrative tasks and customize the environment; Connect data across its client’s marketing ecosystems to their individual customers to orchestrate a relevant experience Analyze the performance of their efforts; Implement security policies that align with business best practices.   Finally, DX Marketing was up and running on Oracle Cloud in just two to three weeks , compared to the four months the company had expected DX Marketing is now able to analyze customer and prospects down to the precise, geospatial location and unique interests when they are online searching for products and services using their huge offline and online customer records to serve highly targeted, personalized, and relevant digital display ads to specific devices. For a residential real estate developer, they are have achieved a 1200% ROI from previous general display campaigns, attributed to serving more relevant ads at more relevant times. They experiencing a 400% lift in conversions over the control group,/a status quo of an already sophisticated, data-driven marketing approach. They have decreased the cost of customer acquisition by over half - the lowest acquisition cost of any other prospecting efforts. They have strategically transformed their campaign thinking from the traditional Key Performance Indicator of Clicks to Conversions.
• #30: ある都市警察の取り組み もともと、警察官の経験と勘に頼って警官を配備していたが、効果的に犯罪を防止できなかった そこで、あらゆるデータから犯罪予測モデルを構築し、犯罪予測モデルを構築、犯罪発生確率の予測を実現しました。 しかも、予測分析モデルを業務アプリケーションに組み込むことで、一部の分析官だけでなく、現場の警察官も活用可能に 現場の警察官が今日、今何をすべきか把握できる
• #31: 予測モデルを地図と重ね合わせて可視化 地区ごとに犯罪発生確率を予測 犯罪発生確率が高い地区は赤く表示
• #32: 過去の犯罪情報、911のコール情報など、あらゆる情報をDB上に集約 数人の分析官がモデルを構築し、その後はデータを与えて機械学習によって自らモデルを改善 そのように構築した予測分析モデルを業務アプリケーションに組み込み 一部の分析官だけでなく、現場の配備担当者も活用可能に このように誰もが予測分析モデルを活用できるようにしたが、やはり予測分析モデルは構築しなければならない しかし、前半、大橋がご説明したように、データサイエンティストの確保は容易ではない。 この課題を解消するため、弊社製品はR言語を活用してモデルを構築する