||Introduction to KNIME
KNIME Analytics Platform is an open source software for working with all kinds of data. It uses visual workflows that are created with an intuitive, drag and drop style graphical interface, without the need for coding.
One can selectively execute some or all nodes in a workflow, and inspect the results using interactive widgets and views. KNIME comes with a rich set of nodes and can easily be extended with code from R, python or Java.
Some key features:
- Data blending
KNIME can import data from simple text formats like CSV, XLS, JSON or similar, unstructured data types like images, documents, networks, or time series data as well as all major DBMS or cloud storage.
- Data shaping
Data cleaning can be done through normalisation, data type conversion, and missing/out-of-range value handling. Data can be aggregated, sorted, filtered and joined in any way imaginable. Included functions range from descriptive statistics and statistical hypothesis validation to dimensions reduction and correlation analysis.
- Machine learning
You can build machine learning models, optimize their performance and validate them against metrics or test sets.
About the webinar
The webinar gives an introduction to KNIME and focuses on data blending and shaping.
Magnus Pfeffer is professor for information management at Stuttgart Media University, Germany. His research interests include linked data, metadata integration, information retrieval in heterogeneous datasets and machine learning.
Dr. Kai Eckert is professor for web-based information systems at Stuttgart Media University, Germany. His research interests include linked data, data integration and enrichment, knowledge organization and data provenance.