Oracle introduced a new set of capabilities in Oracle Analytics Cloud. With new innovations, customers can now experience interactions with their data via maps, visual market basket analysis, and mobile devices to more quickly identify patterns and relationships.
New capabilities include:
Oracle Analytics Cloud enhancements include explainable machine learning, data preparation for transforming customer-specific data into quality information, built-in text analytics, affinity analysis, custom reference knowledge, graph analytics, custom map analytics, natural language queries, and narratives, as well as a new mobile app.
- Explainable Machine Learning: Any user can now see simple explanations of the factors that influenced a machine learning model to predict a certain outcome. In addition, they can interact with a model, adjusting factors to fine-tune the results. For example, of all the factors that influenced the denial of a bank loan application, users can quickly see which were the most determinant and why?
- Automated Data Preparation: A data profiling engine samples and scans data to identify and proactively prompt users about potential data quality issues, like automatically suggesting the obfuscation of sensitive credit card information or social security number. It can enrich zip codes with city, population, income, ethnicity, and payment data to provide more in-depth location analysis. Users can further enrich data by uploading more business-specific data, such as sales regions, delivery zones, or product categories.
- Text Analytics: Text analytics enables you to extract words from unstructured data, count them, visualize the results, and then join that analysis with your original data so you can drill into any level of detail. For example, sentiment analysis uses text analytics to determine whether comments are negative, positive, or neutral, enabling users to understand how their brand is perceived or how a product launch is performing based on text in surveys or social media.
- Affinity Analysis: Discover relationships in your data by identifying sets of items that often appear together. This data mining technique is also known as association rule learning. A common and useful application of it is market basket analysis in consumer goods or retail banking, where users can obtain the probability of different products being purchased together to make marketing decisions. When developing promotions, retailers often look at popular combinations to develop their strategies for increasing product sales. For example, shoppers who buy cereal often also buy milk at the same time. Understanding this co-occurrence of items in a collection helps retailers better manage store layout, coupon offers, and cross-selling, and is valuable for direct marketing, sales promotions, and discovering business trends.
- Graph Analytics: Graph analytics show data relationships visually, such as how people and transactions are connected or the shortest distance between two hubs in a network. Using Oracle Analytics Cloud, anyone can easily analyze graph data in the Autonomous Data Warehouse. This has powerful applications in a variety of various domains, from marketing and social media to security and compliance. For example, pathfinding lets a user find the shortest path between two points; another common use of graph analytics is for ranking and measuring the importance of website pages.
- Custom Map Analytics: Map analytics gives users the ability to apply custom images as map backgrounds and create map layers to enhance data visualizations. For example, doctors can visualize data on an image of the human body to identify areas that require attention and visually track the impact of medication or other treatments. Maps can be loaded into OAC or hosted on a web server as a dynamic background using the Web Map Service (WMS) protocol and XYZ tile layers. This enables customers to use map information they might not have access to in their enterprise, such as weather and building schematics, and easily present it with their business data.
- New Oracle Analytics Mobile App: The new Oracle Analytics mobile app lets users find data quickly and easily, all with a consistent user experience across Oracle Analytics Cloud and the app. It lets users interact with data visualizations, explore dashboards, and share information across teams for further collaboration. Users can also listen to natural-language generated audio narratives of the most salient points from reports, dashboards, and visualizations.
- Natural Language Processing: Oracle Analytics Cloud allows users to query their data in natural language using a simple search-like experience, using text or voice, and obtain spoken narratives of the results. It supports 28 different languages and various language constructs such as synonyms, abbreviations, dynamic filters, and calculations. Users can type, text, or speak aloud to ask business questions, such as, “what’s our employee churn this month?” and get an employee attrition dashboard in return. Oracle Analytics Cloud not only accepts natural language as input, but it also outputs natural language narratives that explain the results of the query. It has an embedded natural-language generation engine that understands the context of the data a user is looking at and automatically updates the narrative as the user adds data, changes filters, or otherwise changes the context as in a typical data discovery and analysis process.