Any machine learning model is based on data. Normally, huge amounts of data a model will learn from. Labels are annotations given to pieces of raw data that are needed for their identification. Labels identify whether an image depicts a daisy or a building, which words were said in a record, or whether item packaging has defects. Data labeling is necessary for a number of applications such as computer vision, natural language processing, and speech recognition.
Big data refers to vast and complicated data sets that are created and collected in real time from multiple sources.
Speech recognition is the technology that enables computers to recognize and transcribe human speech so that people can interact with computers using spoken language.