Wednesday 12 March 2025
Researchers have been studying user preferences for top-k queries, a type of query that retrieves the top-N items in a database based on certain criteria. Top-k queries are commonly used in e-commerce and other applications where users need to quickly find the best options among many alternatives.
The study, published recently, aimed to investigate how users prefer their top-k results when given two different types of queries: linear top-k query and directional top-k query. Linear top-k query is a traditional approach that simply ranks items based on their attribute values, while directional top-k query takes into account the spatial relationship between data points and a user-defined preference line.
The researchers conducted an experiment with 106 participants who were asked to evaluate the results of both types of queries for two specific topics: used cars and football players. The participants were also asked to rate their preferences for each query type and provide feedback on the relevance of the results.
The study found that users tend to prefer directional top-k queries in lifestyle-related domains like e-commerce, where balanced results align with user goals. In contrast, linear top-k queries are more popular in specialized domains like sports, where users may be looking for extreme attribute values like high scores or low prices.
Interestingly, the researchers found that demographic factors like age and expertise had little impact on user preferences in lifestyle-related domains, while they played a significant role in specialized domains. For example, knowledgeable users tended to prefer linear top-k queries in the football players topic, where they could find more extreme attribute values like high scores or low salaries.
The study’s findings have implications for database systems and search engines that rely on top-k queries to provide users with relevant results. By incorporating directional top-k queries into their systems, these applications can improve user satisfaction and provide more balanced results in lifestyle-related domains.
One potential limitation of the study is its focus on a small sample size of 106 participants, which may not be representative of the broader population. However, the researchers note that this was an intentional choice to ensure that participants were thoroughly familiar with the topics and queries being tested.
Overall, the study provides valuable insights into how users prefer their top-k results in different domains and contexts. By understanding these preferences, database systems and search engines can improve their performance and provide better user experiences.
Cite this article: “User Preferences for Top-K Queries in Different Domains”, The Science Archive, 2025.
Top-K Queries, User Preferences, Linear Top-K Query, Directional Top-K Query, E-Commerce, Database Systems, Search Engines, Lifestyle-Related Domains, Specialized Domains, Demographic Factors
