Language technologies are inevitable for adding value to Big Data. They are needed for extracting information and knowledge from textual documents and social media, and for making them accessible across language borders. But they are also required for searching, linking and documenting non linguistic, e.g. numerical or multi media data, through effective metadata. And they are the core of intuitive semantic interfaces to the human consumer of the end products, e.g., through question answering, report generation and summarization. The proposed joint panel by META-NET, LT Innovate and Multilingual Web is a call for action: to further explore the impact of language technologies on big data and to develop a plan - for leveraging big data to big knowledge.