Designing an Article Retrieval Interface with Journalist Usability in Mind -Support the Validation Process-
Chica Takeda (AY 2011)
Today, newspaper circulation in Japan exceeds 50 million copies per day, and nearly all households subscribe to a newspaper. Journalists who write articles for newspapers use their own article search systems and other tools to gather information as they work on their articles. There is Simon's research on journalists' information-seeking behavior; Simon defined four tasks for which journalists use information-seeking: identifying news, promoting understanding, finding content, and accumulating information.However, there are three major roles of journalists: reporter, editor, and organizer (Kanmon), but their relationship to Gimon's four tasks has not been clarified.Therefore, the first objective of this study was to clarify the relationship between Simon's four tasks and the three roles of journalists. We decided to design a search interface that takes the tasks and roles into account.
For the first objective, we conducted a questionnaire survey among the journalist positions, editors, and the organization department (Kanmon) of a certain news agency.As a result, there was no noticeable difference in the weight of the four tasks among the three roles, although there was some difference in the weight of the four tasks among the three roles.On the other hand, all roles made verification of novelty one of the search goals. The survey results showed that all three roles had the verification of novelty as one of their search objectives.Therefore, we focused on supporting the verification of article content in particular as the second objective.
This study used user-centered design (UCD) for interface design, a method that considers the user at every step of product development. Specifically, we conducted questionnaires and interviews with journalists, created scenarios based on these interviews, and extracted needs for an article search system. Ideas for the needs were developed through the creation of paper prototypes.
Through a series of interviews, it was found that the journalists targeted in this study perform five main types of validation: validation of phrases, definitions, accuracy of explanations, dates and times, proper nouns, etc., validation of subjectivity, and validation of novelty. Based on this, we created a mockup of an interface that could perform the five verifications. For the novelty check, we thought that some new information could be found by comparing similar articles.To make it easier to find new information, we considered a design that highlights duplicate keywords and dates, and created a mockup.
A subject experiment was conducted to verify the effectiveness of the highlighted articles in confirming novelty. The results showed that there was no particular effect of highlighted articles on both work time and percentage of correct answers, but the questionnaire gave a good impression of the highlighted articles.Thus, the novelty check was Further research is warranted into the effective support of article review work, including novelty checking.
Future directions include the design and development of a more effective support system for verification work and its evaluation.
(Translated by DeepL)