In this paper the properties of big data are described, with a deeper insight into veracity, value, and variability. A brief overview of all properties and some examples of big data sources are provided. Data veracity is examined through various data quality attributes, and a particular focus is given to the challenges of analysing language and human expression. Some business methods for improving and preserving data quality are mentioned, as well as other automated methods. Various examples of extracting value from big data are detailed. Tools and methods for data processing, including data clustering, regression and predictive analysis, and classification, are presented, with an emphasis on machine learning. Natural language processing and its components are specifically explored. Some platforms for big data analysis are also mentioned. Variability is described with examples of variable data and models. Adaptive models based on concept drift are further explored. Data linkage and integration issues from new sources are discussed, along with examples of algorithms for automatic problem-solving. In the practical part, some data processing methods are demonstrated using social media data as an example. Data format, collection, and preprocessing are described in detail. Veracity, or rather credibility of data is examined using a naive Bayes classifier, categorizing the data into rumours and verifiable news. Value extraction is demonstrated by analyzing public sentiment on topics related to observed events, while variability is explored by finding the causes of sentiment changes related to selected topics, organizations, and individuals.