Sonny, Cerca! Evaluating the Impact of Using a Vocal Assistant to Search at School

Abstract

Children struggle with translating their information needs into effective queries to initiate the search process. In this paper, we explore the degree to which the use of a Vocal Assistant (VA) as an intermediary between a child and a search engine can ease query formulation and foster completion of successful searches. We also examine the potential influence VA can have on the search process when compared to a traditional keyboard-driven approach. This comparison motivates the second contribution of our work, an evaluation framework that covers 4 dimensions: (1) a new search strategy (VA) for (2) a specific user group (children) given (3) a particular task (answering questions) in (4) a defined environment (school). The proposed framework can be adopted by the research community to conduct comprehensive assessments of search systems given new interaction methods, user groups, contexts, and tasks.

To Infinity and Beyond! Accessibility is the Future for Kids' Search Engines

“…When looking at results retrieved in response to children's queries, Anuyah et al [2] and Bilal and Huang [9] investigate the reading level of SERP and whether they align with a child's reading level. When considering the relevance of retrieved resources, children tend to explore search engine result pages (SERP) sequentially, i.e., from top to bottom [26,32], as opposed to deciding relevance through scanning SERP snippets and clicking on a given result. This prompted researchers to investigate ways to both filter and re-rank resources so that those more relevant and suitable for children make their way to the top of the SERP [7,13,18,20,38,40,48].…”

An empirical analysis of search engines’ response to web search queries associated with the classroom setting

“…Although children frequently turn to SEs, research has shown that they often experience difficulties in finding resources that satisfy their information needs (Bilal and Gwizdka, 2018; Bilal and Ellis, 2011; Druin et al , 2009; Gossen, 2016). Prior works suggest that this difficulty is often the result of children’s insufficient digital literacy skills (Bilal and Gwizdka, 2018; Hague and Payton, 2011), for example, lack of ability in identifying credible resources when using SEs, in addition to varied degrees of prior experience (Landoni et al , 2019) and domain knowledge (Han, 2017; Yamin et al , 2013). Moreover, children favor resources on the first SERP and barely notice that other SERPs exist (Duarte Torres et al , 2010b; Gwizdka and Bilal, 2017).…”

Does your robot know? Enhancing children's information retrieval through spoken conversation with responsible robots

“…While voice agents have become available, which children also use to search for information [9,[20][21][22], the interaction that these voice agents provide is of a simple question-answer style, and not conversational. This limits children in that they cannot ask follow up questions like they often assume [16,22], have to include context into complex statements [22,33], and do not get assisted with suggestions or clarifying questions [5,33].…”

“…Inspired by Landoni et al [16], we propose to aid children in searching and exploring (multimedia) information through spoken conversational agents. Our target group is children of 10-12 years old.…”