用户信息搜索中的学习行为及过程探究

Characterizing and Exploring Users’ Learning Behaviors and Process during Information Search

用户信息搜寻的目的是为了学习新知识或完成某项任务,以往研究中学者主要关注搜索行为,但是对于用户如何使用搜索到的信息进行知识积累以完成学习任务的相关研究仍比较缺乏。对用户学习行为和过程加以探索,深入理解用户信息搜索过程中知识学习的发生、演进特征,有助于我们更好地设计支持用户学习任务下的信息搜索系统。本研究采用实验法,被试者在实验室环境下完成四个与学习相关的搜索任务,在搜索过程中要求被试者在电脑的记事本文件中记录和整理自己认为对完成任务有用的信息或答案。这种设计可使用户的学习过程通过其信息记录行为外显化,进而通过分析用户的信息记录行为特征,揭示用户的学习行为特征。在数据分析中,依据不同阶段用户记录完成度的推进情况,将他们的记录策略分为三种不同的模式:早期记录型、平均记录型、后期记录型。研究还探索了影响用户记录策略的因素,发现用户的自信程度、话题熟悉度、任务类型以及任务难度都会显著影响用户采取何种记录策略。在不同记录策略与搜索交互、任务产出的关系分析中发现,平均记录策略作为记录行为随搜索过程逐步推进的一般模式,学习产出的质量相对更好,因此相对而言更有效率。这些发现有助于深入理解用户的信息使用行为和学习发生过程,启发支持学习相关搜索的信息系统的设计与优化。

Users search information in order to learn new things or to complete a task.Previous studies mainly focused on users’information search behavior,and it has not been extensively examined how users would use the information they have obtained for accumulating knowledge in order to accomplish learning-related tasks.In this study,we conducted a controlled user experiment,assigned four learning-related tasks for participants to search and asked them to write down notes on the computer during search.Users’learning process was manifested through their writing behavior,so that users’learning behavior characteristics were revealed by analyzing their writing behavior characteristics.Based on their writing process along with searching,their search sessions were divided into three types:early writing strategy,average writing strategy and late writing strategy.After characterizing three writing strategies,we further examined which factors were related to searchers’writing strategies.It was found that searchers’writing strategies were influenced by users’pre-confidence,topic familiarity,type and difficulty of the task.The examination of search interactions and task outcome indicated that"average writing strategy"was the normal pattern and also the most efficient strategy with best quality outcome.These findings help understand searchers’information use behavior and learning process while searching,and have implications on search system design to support learning-related search task accomplishment.