Business Analytics using Semantics
The purpose of Semantic-enabled Business Analytics
Do you have all the information required for decision making? Are your IT systems capable to answer really important questions, such as:
- Where can I find business opportunities?
- How are my customers making decisions?
- How to improve my products to better fit the market?
- Why my business offers are sometimes rejected?
- How to optimize my manufacturing plans?
It is obvious that you need to integrate information from various data sources to answer any of these questions. These sources may lay within your corporation (CRM, ERP etc), may be external (news feeds, purchase announces), or may be hidden in minds of your personnel.
You need various data sets to answer various questions - it means that it is impossible to build a Data Warehouse once, and use it in a long period of time. The life makes new questions every day, and to answer it we need to involve new information types and sources.
This problem can be resolved only using Semantic Technologies.
- Semantics rejects building databases with strictly defines structure. Instead, it builds a flexible information model, which can be enriched and extended at any moment.
- Semantic information model unifies data from various sources, expressing it using common dictionary (ontology). It transforms data to the knowledge. This knowledge becomes a trustable source for business decision making.
- Using semantic information model, it is easy to build digital optimization methods for various production, exploration and maintenance tasks.
We encourage you to request the demo access to our Knowledge Extraction System prototype, to familiarize yourself with the power of Semantic-enabled analytics.
How Semantic Analytics differs from BI and OLAP?
Contemporary analytics instuments (BI and OLAP) are mostly focused on the quiantitive, statistical analysis of business data. All such applications are using strictly structured data, sets of parameters and KPI. It is hard to analyse qualitive relationships using such methods (for example, how some factors are affecting result). Semantic analytics, built using Open World concept, is offering new tools for resolving such kind of tasks.
The easiest way to illustrate differences between semantics and other analytical methods is the Facebook Graph Search. This interface, built into the social network, is able to answer questions like "Which restaurants are liked by my friends?", or "In which cities does my relatives live?" It is obvious that no one full-text search system can answer such questions. Relational DB-based systems (like BI and OLAP) could answer such questions only if they fit their data model and user interface. The set of entities and properties in such systems is strictly limited by the database structure. Semantic analytics allow to easily add new entities to the model, not changing physical data storage schema, nor user interface or other parts of the model.
"Business Semantics" solutions for semantic analytics
We have created a demonstration prototype, intended to show new features of such kind of knowledge extraction software. It uses a small piece of NPD (Norwegian Petroleum Directorate) data to demonstrate the power of querying semantic data model. This prototype allows to get familiar with Semantic queries building.
Ontorion Semantic Framework by Cognitum, also included in our product portfolio, is also offering interesting features of Natural Language-based knowledge extraction.