Natural Language Understanding (NLU) are systems where users can interact with the computer via natural human language. NLU programs understand the meaning of questions and complete sentences and can interpret and answer them.
NLU systems go far beyond speech recognition and automatic language recognition systems. They understand the semantic relationships, the meaning and the context of the statement, and can draw conclusions from them. In contrast, pure speech recognition systems, such as those used in IVR systems or in voice portals, work with keywords, which they tag and filter out from spoken texts and compare with stored keywords. Such speech recognition systems work with Part-of-Speech( PoS), a technique with which words are tagged and assigned to a word category, furthermore with Natural Language Processing( NLP) and intelligent dialog components that can be linked as word combinations.
NLU systems can be seen as a precursor to artificial intelligence, supporting direct human-computer interaction. This allows them to issue non-formalized commands to machines in human language and communicate with computers as if they were humans. Understanding includes sentences not pronounced correctly or spoken in dialect. If instructions or statements are not understood, then NLU systems can make dedicated inquiries.