Tools for supporting group decision making

Recommender Systems in Requirements Engineering

Group decision making may be hard, but there are tools to ease the situation.

On many occasions, people make decisions not on their own but as a part of a group of people. A major issue in group decision making is how to deal with conflicting preferences of group members.

How is it possible to resolve the conflict and produce a single recommendation to a group?

There already exist some platforms or group decision support environments to ease the situation on some occasions. These incude, for example, Doodle and Vern used for coordinating meeting times. In addition, there is Smartocracy providing support for voting scenarios, and Dotmocracy offering a method for collecting and visualising group preferences.

Some of the researchers in the OpenReq team have a history of studying group decision making. They have tested such tools as Choicla. Choicla offers the possibility for a flexible definition of domain-independent decision scenarios. It supports the determination of recommendations for groups. It also includes mechanisms to counteract decision biases.

In private settings Choicla may, for instance, help in selecting a restaurant or a hotel for a holiday trip with friends. On the other hand, in industrial settings it may well be used to select a conference location or evaluate project proposals.

In an empirical study that researchers in the OpenReq team conducted, groups of students had to make a decision about exam topics. The groups received a varied amount of recommendations from other group members. The results of the study show, for instance, that increased recommendation diversity leads to an increased degree of information exchange among group members.

OpenReq team develops new tools

One of the aims of the OpenReq team is to further develop group decision making tools and technologies for more complex products and services. The team is on its way to develop a tool called the OpenReq Prototype which is a release planning component.

Studies on existing tools already show encouraging results. Conflicting preferences can be coordinated and resolved even in more complex settings.