Collaborative search for places of interest

Finding a place of interest (e.g., a restaurant, hotel, or attraction) is often related to a group information need. There is a need to understand more details about the collaboration in such search. In this thesis, we aim to study the group search for places in terms of significance, information seeking behaviors, group activities, settings, and issues. We also aim to explore the use of a system setup for supporting this kind of collaborative search. To achieve these goals, we conducted two survey questionnaires and a user study. Below are some of the findings.

Group search for places was highly prevalent, and in most cases involved active participation by multiple people. It occurred in a range of settings with group sizes typically being 2 or 3. Participants were co-located in many cases and commonly shared information via messaging even while together. Many searches stretched over multiple sessions and most of the time over multiple days.

The interaction between participants was dominated by the discussion of search results while being less about planning the task specifications. Looking for more information about places shared by others was highly common, which helped participants spot helpful information and discover useful websites. Although familiar to the majority, dividing the search and discussing query terms were not quite frequent strategies. The disagreement between group members was the most common inconvenience in the group search for places. Deciding on places can be associated with uncertainty and confusion between group members, in addition to the physical and cognitive load of negotiating and searching for alternative places to resolve disagreements. The lack of feedback about shared information from other participants and the difficulty to refind shared information were also common issues.

In the user study, group members influenced the views and work of each other through negotiation and sharing of content. We also found collaboration similarities between the groups. A member in each group took a leading role in the main task discussions, while another was less active than other members in these discussions. Other members were somewhere in the middle. The participants who were the most active in discussing the task were also active in the browsing but browsed the lowest number of pages shared by other members. We presented a representation of the group information seeking process, which connects the typical information seeking activities with their relevant group interactions and the resulting group outcomes. We also suggested a multilevel perspective on group information seeking that distinguishes between the individual information seeking process, group dynamics, and contextual factors.

We investigated the use of a system setup consisting of a webpage for sharing places as photo posts along with a map and a chatting application. The participants’ use and feedback about sharing places using that setup was positive. However, the chat communication was cluttered and some messages were ambiguous; some participants missed parts of the communicated information. The participants also highlighted issues in the system and provided useful suggestions to improve it. Based on the findings, we proposed a user interface design for collaborative search for places.