Big data refers to datasets that exceed the competence of commonly used IT systems in terms of processing space and/or time. Traditionally, massive data are mostly produced in scientific fields such as astronomy, meteorology, genomics physics, biology, and environmental research. Due to the rapid development of IT technology and the consequent decrease of cost on collecting and storing data, big data has been generated from almost every industry and sector as well as governmental department, including retail, finance, banking, security, audit, electric power, healthcare, to name a few. Recently, big data over the Web (big Web data for short), which includes all the context data, such as, user generated contents, browser/search log data, deep web data, etc., have attracted extensive interests, as these context data and their analyses help us to understand what is happening in real life. This can help to enable new ways for improving user experience by providing more accurate predictions and recommendations thus creating a personalized smarter internet. Currently, big data is often on the order of petabytes and even exabytes. However, big data has become bigger and bigger not only in its size, but also in its growth rate and variety. The volume of big data often grows exponentially or even in rates that overwhelm the well-known Moore's Law. Meanwhile, big data has been extended from traditional structured data into semi-structured and completely unstructured data of different types, such as text, image, audio, video, click streams, log files, etc. Moreover, big data is often internally interconnected and thus form complex data/information networks. Although big data can offer us unprecedented opportunities, they also pose many grand challenges. Due to the massive volume and inherent complexity, it is extremely difficult to store, aggregate, manage, and analyze big data and finally mine valuable information/knowledge from the complex data/information networks. Therefore, in the presence of big data, the models, algorithms and methods for traditional data mining become no longer effective and efficient. For instance, similarity learning, upon which various similarity-based tasks (e.g., ranking and clustering) can be launched, is extremely challenging for real applications with big data due to their typical features such as the data being heterogeneous, time-evolving, sparse and noisy. On the other hand, some data is generated exponentially or super-exponentially in a streaming manner. Therefore, how to carry out real-time analysis on, and deep mining and understanding from big data so as to obtain dynamical and incremental information/knowledge, is another grand challenge. In general, at the era of big data, it is expected to develop new models, algorithms, methods, and even paradigms for mining, analyzing, and understanding big data. This workshop aims to provide a networking venue that will bring together scientists, researchers, professionals, and practitioners from both industry and academia and from different disciplines (including computer science, social science, network science, etc.) to exchange ideas, discuss solutions, share experiences, promote collaborations, and report state-of-the-art research results and technological innovations on various aspects of mining and understanding from big data.