AI-901 - Identify AI concepts and capabilities (42% of the exam) - Section 1.3
Identify scenarios for common AI workloads and the techniques within text analysis, speech, computer vision, and information extraction.
Matching a scenario to the right AI workload, including generative and agentic AI, text analysis, speech, computer vision, and information extraction; keeping the text-analysis techniques distinct (keyword extraction, entity detection, sentiment analysis, and summarization). Agentic AI is distinct from generative AI (an agent plans and takes actions towards a goal), and information extraction is a distinct workload from general computer vision.
generative AIagentic AItext analysiskeyword extractionentity detectionsentiment analysissummarizationspeech recognition and synthesiscomputer vision and image generationinformation extraction
Practice question for this objective
lock_openFree sampleIdentify AI concepts and capabilitiesmedium
A subscription business collects thousands of free-text customer reviews each week and wants an automated way to classify each review as expressing a positive, negative, or neutral opinion, so the support team can prioritise unhappy customers first. Which text analysis technique fits this requirement?
AKeyword extraction, which surfaces the main talking points and terms mentioned across the review text
BEntity detection, which locates the named people, organisations, and places referred to in the review
CSentiment analysis, which scores each review by the positive, negative, or neutral opinion it expressescheck_circle Correct
DSummarization, which condenses each review down into a shorter version of its overall content
Sentiment analysis classifies text by opinion polarity, distinguishing it from keyword extraction, entity detection, and summarization. Sentiment analysis is the text analysis technique that assigns a polarity, positive, negative, or neutral, to a passage based on the opinion it conveys, which is precisely what prioritising dissatisfied reviewers requires.
Why A is wrong: Keyword extraction reads the same review text, which makes it tempting, but it only returns the salient terms and never judges whether the opinion expressed is favourable or not.
Why B is wrong: Entity detection identifies typed things such as brand or product names, so it looks plausible for review text, but it reports what is mentioned rather than the writer's opinion about it.
Why C is correct: Sentiment analysis evaluates the emotional tone of a piece of text and returns a positive, negative, or neutral classification, which is exactly what is needed to flag dissatisfied customers.
Why D is wrong: Summarization shortens the text while keeping its meaning, which sounds useful, but a shorter review is still not labelled by opinion polarity, so the team could not rank unhappy customers.
Exam traps in Identify AI concepts and capabilities
Answers that look right on this material and are not. Each one is a distractor from a different question in the AI-901 bank for this domain.
Keyword extraction, which surfaces the most salient terms in a document but does not judge the writer's opinion.
errorWhy it is wrong: Keywords help summarise what a review is about, which makes this tempting, but the technique ranks terms and never assigns a positive or negative judgement.
Generative AI, because a model produces fluent written replies to the employee's original report
errorWhy it is wrong: Generative AI is tempting because such a system may phrase replies in natural language, but generation alone only creates content and does not plan a sequence and call tools to act on the world.
Sentiment analysis, which returns whether the article reads as positive, negative, or neutral in tone
errorWhy it is wrong: Sentiment analysis is also a text analysis task, which makes it tempting, but it reports the emotional polarity of the article rather than the named things mentioned inside it.