Artificial Curation - IELTS Reading Answers & Explanations
From Collins Practice Tests For IELTS 3 Academic Reading Test 4 · Part 3 · Questions 30–40
Reading Passage
You should spend about 20 minutes on Questions 30-40, which are based on Reading Passage 3 below.
Artificial curation
We all know that too much information can be a bad thing - this is as true in daily life as it is in business. Filtering useful from useless information has become a growing problem, bringing confusion with it, but this is where data curation can help. Curating data involves finding and displaying patterns in large volumes of disconnected and messy data to create meaningful information and suggestions for the end user. The process of data collection to inform business and consumer choice has developed from collecting information via questionnaires and interviews to digitalising information and using technology to gather and interpret data. The latter requires huge databases for computer algorithms (sets of computer instructions) to search and find patterns in order to predict what choices we might make. The use of algorithms is called Al curation, and from shopping to social media, it is part of our lives.
Al curation involves designing computer algorithms that work with large amounts of data. The data is gathered from people's past internet use, for example, searches, purchases, likes and bookmarks. The algorithm looks for patterns in this historical data and uses these patterns to predict the user's choices. The patterns and predictions help the algorithm search and sort through the huge volume of information on the internet and present items that the user has previously looked for or liked or bought. For example, if a person has searched for a particular product online, the algorithm may make suggestions for other products or websites the person might be interested in. If the consumer goes on to click through to the website or to buy something that has been recommended, then the algorithm has found a winning pattern for that individual.
Algorithms make it possible to collect data about a target audience and consequently, they determine which adverts we see as we browse the internet as well as which news stories are shown to us. The purpose of this is to increase our engagement with a particular company or website and thus generate revenue. Al curation has other benefits too. In the recent past, when we wanted information about the news, we went to a news site and if we were interested in sports news, for example, we searched for sports news within that news site. With Al curation, based on our searches and our likes, we have the sports news brought directly to us. Users see what they are mostly interested in and, conversely, see less of what they don't want to. They can therefore build a relationship with websites they frequently visit as content and products are personalised for them, building loyalty and trust. Finally, they are able to focus on certain information and build up specialist knowledge about the things they are interested in. They can even connect with other users who are interested in the same subject and form communities based on that particular topic.
Although algorithms are good to some extent at curating virtual information for us and putting forward suggestions, some companies are moving away from Al and back to human recommendations. Humans have always played a role in some areas of internet curation, particularly news or stories where people have to moderate or make a choice about what is suitable or not for an audience. There have been several cases where algorithms have spread stories that were partially or wholly untrue or presented content that was not suitable for younger people. Some social media companies are bringing back human editors because algorithms cannot distinguish between stories that have on-the- record sources and stories that are simply made up. The other thing a recommendation algorithm can't do is to tell you why it is desirable to use a product – what makes it great or different or better than a rival product. Neither can it give its recommendation the human touch, and people usually prefer recommendations from other people. The problem with human curation, however, is that there just isn't enough of it to deal with all the information, particularly when algorithms are cheaper and much more efficient.
It seems that the best way forward is a solution that uses both Al and human curation. One news company that works in this way starts by asking its users to pick articles for others to read. These stories form the basis for algorithms to work with. The algorithms then gather similar stories for the next stage. When the algorithm has compiled stories from various internet sources, a human editor fine-tunes the selection for the audience. For example, if the Al curator is asked for stories about famous people, it may deliver sensational or even misleading stories because these are the ones that have received the most clicks. At this point, the human curator intervenes and deletes inappropriate content in favour of better stories. It seems that putting a human curator alongside Al can give a better quality of curation and create trust in the recommendations as well as a personal connection that isn't offered by an algorithm alone.
Questions
Questions 30–32 Sentence Completion
Complete the sentences below.
Choose ONE WORD ONLY from the passage for each answer.
- The amount of data that is available to companies is too great to be useful and results in 30
- Al curation means that a computer program looks for and shows 31 in large amounts of data.
- Technology is applied to digital information to collect and 32 the data
Questions 33–34 Multiple Choice (Two Answers)
Choose TWO letters, A-E.
Which TWO things does an algorithm do?
Questions 35–36 Multiple Choice (Two Answers)
Choose TWO letters, A-E.
Which TWO things can't an Al algorithm do?
Questions 37–40 Flow Chart Completion
Complete the flow-chart below.
Choose ONE WORD ONLY from the passage for each answer.
The news company's readers select articles to recommend to other users.
↓
The news items become the 37 algorithms use to collect comparable items.
↓
A human editor 38 the stories collected by the algorithm.
↓
The human editor 39 content that is not suitable and replaces it.
↓
Selection by Al and humans can give a better standard of 40
Answers & Explanations Summary
| # | Answer | Evidence | Explanation |
|---|---|---|---|
| Q30 | confusion | Filtering useful from useless information has become a growing problem, bringing confusion with it | Excerpt/Passage Explanation: The passage explains that when there is too much data, it is hard to separate what is good from what is bad, and this situation creates a lack of clarity for people. Answer Explanation: The answer "confusion" means a state where things are not clear, making it difficult for people or businesses to understand or make decisions. Reason For Correctness: The correct answer is supported by the text which states that having a large volume of information makes it hard to filter what is useful. This difficulty in sorting information is described as a "growing problem" that brings "confusion" to the user. Keywords to notice are "filtering," "problem," and "bringing." |
| Q31 | patterns | Curating data involves finding and displaying patterns in large volumes of disconnected and messy data to create meaningful information and suggestions for the end user | Excerpt/Passage Explanation: The passage says that the work of organizing data includes looking for and showing repeated behaviors or designs (patterns) within a lot of messy information to help the person using the computer. Answer Explanation: The answer "patterns" refers to regular or repeated ways in which something happens or is done. Reason For Correctness: The correct answer is "patterns" because the text explains that AI curation involves using computer instructions (algorithms) to scan through massive amounts of information. The goal of this process is to identify and show repeating behaviors or connected pieces of information, which the text specifically calls "patterns," to help make predictions or suggestions. |
| Q32 | interpret | The process of data collection to inform business and consumer choice has developed from collecting information via questionnaires and interviews to digitalising information and using technology to gather and interpret data | Excerpt/Passage Explanation: The passage explains that business data collection has moved from simple surveys to using technology to find and explain the meaning of digital information. Answer Explanation: The answer "interpret" means to explain the meaning of something or to translate information into a form that is easy to understand. Reason For Correctness: The correct answer is "interpret" because the passage explains how the methods for handling data have changed over time. It states that current methods use technology not only to collect (gather) data but also to make sense of it. In the sentence from the text, "gather" is a synonym for "collect," which matches the structure of the question asking what technology does to the data alongside collecting it. |
| Q33 | — | — | |
| Q34 | A / B | The data is gathered from people's past internet use, for example, searches, purchases, likes and bookmarks. The algorithm looks for patterns in this historical data and uses these patterns to predict the user's choices Users see what they are mostly interested in and, conversely, see less of what they don't want to. They can therefore build a relationship with websites they frequently visit as content and products are personalised for them, building loyalty and trust |
Excerpt/Passage Explanation: The passage explains that these computer programs gather information about what you did on the internet before. Because they show you things that are chosen just for you, you can develop a strong, positive connection with the websites you visit often. Answer Explanation: The answer identifies two actions performed by computer programs (algorithms): first, they look at a person’s history on the internet to collect information; second, they help people feel a stronger connection or bond with the websites they enjoy using. Reason For Correctness: The correct answer is supported by the text in the second and third paragraphs. Paragraph 2 explains that AI curation uses programs to look at "historical data"—which means things you did in the past, like searches or buying items—to understand what you like. Paragraph 3 then explains how this service creates a better experience for the user. Because the program shows the user things they are interested in, the user can "build a relationship" with those websites and feel "trust" and "loyalty" toward them. |
| Q35 | — | — | |
| Q36 | B / C | There have been several cases where algorithms have spread stories that were partially or wholly untrue or presented content that was not suitable for younger people. Some social media companies are bringing back human editors because algorithms cannot distinguish between stories that have on-the-record sources and stories that are simply made up | Excerpt/Passage Explanation: The passage explains that AI programs have shared fake news and materials that were not safe for children. It also says that companies need human workers because computers cannot see the difference between news from real, official sources and stories that are completely fake. Answer Explanation: The answer means that AI systems cannot tell if a story is from a real source or if it is fake news, and they sometimes show content that is not right or safe for the user. Reason For Correctness: The correct answer is B and C because the passage mentions specific limits of computer programs (algorithms). While AI is fast, it cannot tell the difference between news with real sources and stories that are just made up. Additionally, the text notes that AI has shown content that was 'not suitable' (which means not appropriate) for younger people. The word 'distinguish' in the passage relates to choice B, and 'not suitable' relates to choice C. |
| Q37 | basis | These stories form the basis for algorithms to work with | Excerpt/Passage Explanation: The passage explains that the news stories picked by people are the starting material (the basis) that computer programs use to begin their task. Answer Explanation: The answer refers to the foundation or starting point that something else is built upon. Reason For Correctness: The correct answer is 'basis' because the passage describes a specific process used by a news company to filter information. It mentions that users first choose articles, and these selected items then serve as the 'basis'—or the starting data—for computer algorithms to find similar content. Therefore, the news items chosen by people act as the foundation for the computer's subsequent search. |
| Q38 | fine-tunes | When the algorithm has compiled stories from various internet sources, a human editor fine-tunes the selection for the audience | Excerpt/Passage Explanation: The passage explains that once a computer program has finished collecting news stories from the internet, a person who works as an editor makes small improvements to that list of stories for the readers. Answer Explanation: The answer means to make small, careful changes to something so that it works better or is more accurate for the people using it. Reason For Correctness: The correct answer is found in the final paragraph, which describes a combined process used by a news company. After the computer algorithm gathers many similar news stories, a person—specifically a 'human editor'—performs the next step. The passage explicitly states that this person 'fine-tunes' the stories that the algorithm has collected to ensure they are of high quality for the audience. This matches the flow-chart's description of a human editor taking action on the stories gathered by the algorithm. |
| Q39 | deletes | At this point, the human curator intervenes and deletes inappropriate content in favour of better stories | Excerpt/Passage Explanation: The passage explains that a person called a curator steps in to remove bad or wrong information so that only good stories remain. Answer Explanation: The answer 'deletes' means to remove something from a list or a group because it is not needed or is not good. Reason For Correctness: The correct answer 'deletes' describes the action a human editor takes with content that is not right for the readers. In the flow-chart, this step focuses on what happens to 'content that is not suitable'. The passage uses the word 'deletes' specifically when talking about removing 'inappropriate content', which is a synonym for content that is not suitable. |
| Q40 | curation | It seems that putting a human curator alongside Al can give a better quality of curation and create trust in the recommendations as well as a personal connection that isn't offered by an algorithm alone | Excerpt/Passage Explanation: The passage explains that when humans and computers work together to select information, the final result is much better and more trustworthy than if a computer did it by itself. Answer Explanation: The answer 'curation' refers to the process of carefully choosing, organizing, and sharing information or content so that it is useful for others. Reason For Correctness: The correct answer is 'curation' because the flowchart describes a process where both computer programs (AI) and people work together to choose news stories. The final step of the flowchart asks what becomes better when both are used. According to the text, combining human judgment with AI results in a "better quality of curation." In this context, the word 'standard' in the flowchart is a synonym for 'quality' in the text. |