Natural language processing (NLP) teaches computers to work with human language, and its difficulty comes from ambiguity: the same words can mean different things, and meaning depends on context. A typical NLP pipeline processes text in stages. Tokenization splits text into tokens (words and punctuation). Morphological analysis reduces words to base forms: stemming chops endings by rules (running becomes runn or run), while lemmatization uses a dictionary to reach the true base form (better becomes good). Part-of-speech tagging labels each token as noun, verb, adjective and so on. Parsing builds the sentence structure. Semantic analysis assigns meaning, and pragmatics handles context, such as what "it" refers to in a sentence.
Expert systems are an older, rule-based branch of AI: they capture a human expert's knowledge as rules and reason with it. An expert system has a knowledge base (facts and if-then rules), an inference engine (which applies the rules), a working memory (current facts), an explanation facility (which justifies conclusions) and a user interface. They excel in narrow domains with clear rules, such as medical diagnosis or equipment troubleshooting, and they fail where knowledge is tacit or rules are unclear. Fuzzy logic extends this idea: instead of true or false, a statement can be partly true, with a membership degree between 0 and 1, which suits vague concepts like "tall" or "hot".
Worked example — stemming versus lemmatization. Take the words "running", "better" and "geese". A stemmer applying suffix rules gives roughly "run", "better" and "gees": fast, but "better" and "geese" are wrong as base forms. A lemmatizer consulting a dictionary gives "run", "good" and "goose": correct, but slower and language-dependent. The exam tests exactly this trade-off: stemming is crude and fast, lemmatization is accurate and expensive. Similarly, ambiguity types are tested by example: lexical ambiguity is one word with two meanings ("bank" as river bank or money bank), syntactic ambiguity is one sentence with two structures ("I saw the man with the telescope").