Emotional Intelligence Exercise

Introduction

EI is important, but has been a hard thing to develop. Emotional intelligence (EI) describes a person's ability to perceive, express, use, understand, and manage his/her emotions or feelings. To measure and develop emotional intelligence, psychologists have developed tests of EI, such as Mayer-Salovey Caruso Emotional Intelligence Test (MSCEIT), as well as counseling programs. Although existing approaches for developing people's emotional intelligence are well-designed, they tend to require both trained and experienced counselors, and great expenses. Thus, for people who are not heavily emotionally impaired, fostering EI may be viewed as luxurious and even unnecessary. This is unfortunate because well-developed emotional intelligence has been proposed to be an important component of good social relationships, career, marriage, and life.

EI can be developed by ordinary people’s efforts. I argue that EI can be developed in much easier ways, since everyone has the ability to support others with their emotions. In a conversation with participants A and B, where B tries to make clear his/her own feeling and find a way out and A is to support him/her, I believe both A's and B's EI can be improved over the support process -- by being B (the supported side), participants will be encouraged to find out their own feelings and the best actions to take, whereas being A (the supportive side) will be a training of identifying the partner’s emotions and of providing support. The only prerequisite here is that their attention needs to be focused on the emotional states as how the experts do. Once they can keep fixed to the right focus, their EI will be improved as conversation proceeds.

We can help people concentrate on developing each other’s EI. To this end, I propose to build this EI Exercise system. In the EI Exercise environment, participants can communicate with each other anonymously through the Internet. The system, on the other hand, prompts feedback suggestions based on affective computing techniques, and concentrates the participants' attention on B’s emotional states. The basic idea of this system is that computers are short of understanding context, but humans are experts at it. On the contrary, humans usually fail to keep concentrated on one focus as conversation flows, but computers are good at such precise tasks. By combining the strength of both humans and computers, psychology expertise may be virtually achieved.

We expect, after one after another conversation, the emotional intelligence of A, B, and the system itself, can be enhanced. Specifically, the benefits of EI exercise may include:

1. Helping participants foster their self-awareness ability
   - Knowing how they feel
   - Knowing what they need to feel
2. Helping participants foster their management ability of emotions
   - Knowing what to do
3. Mirroring partner's emotions and behaviors
4. Building a computational semantic network that relates life events with possible personal feelings, such that computers may learn
   - how to recognize people's feelings from words
   - how to respond when people have such feelings
   - what may happen when people have such feelings

The System Interface




The above figure shows how an EI-exercise system may look. This is the window seen by partner A (the supportive side). Three sub-windows are on the left, including, in a top-down order, the conversation log, the typing area, and the system suggestion area. On the right you can see the “Emotion Crayons,” where dozens of feeling terms can be found in a hierarchical directory, such as “upset” and “furious” in the “angry” category, "Feedback Guidelines," which are reminders for better attitudes, and “Partner B Needs to be…,” which is another list containing terms that describe emotional needs. We now introduce the details of each of the six parts.

1. Conversation Log: All sentences during the conversation are shown in this window. The participants will be either partner A or partner B in a conversation, depending on their own willingness, and only the system knows their identities beside themselves.
2. Typing Area: Users can type in sentences in this window, and send them by pressing the “Send” button.
3. System Suggestion Area: This is a pull-down list where the system prompts suggestion for the feedback. The best suggestion (e.g., “How do you feel when you think of such burden?” in the figure) is directly visible, and the others can be seen and selected after the user presses the button with the triangle. And, the selected suggestion will be copied into the typing area for further adjustments when the “suggest” button is pressed. The suggestions are proposed based on the system’s learning result over conversation sessions.
4. Emotion Crayons: Users can select the emotion crayons and label the sentences in the log sub-window as excited, frustrated, and other emotions. With such visual assists, participants will be focused on the emotions, and the system will also be able to record and utilize the relationships between emotions and sentences. The crayon types are originated from the feeling words here: http://eqi.org/fw.htm.
5. Feedback Guidelines: If the user is not sure how to feedback or wish to remind himself/herself a better attitude of viewing things, he/she can click on the feedback guideline button to see this form: http://eqi.org/summary.htm. It helps the participants to develop their emotional intelligence in a more effective way.
6. Emotional Needs: We also list dozens of emotional needs for users to choose to help keep their focus. The emotional needs will also be recorded as one’s personal profile.

Using the system

1. Before every conversation begins, feedback guidelines for developing EI proposed by psychologists are shown to remind the participants not to mis-focus or mis-behave.
2. For each user, after logging into the system and identifying as willing to be A or B, the system finds him/her a partner from all on-line free participants. Two participants are grouped as partners based on the similarities of their personal profile, which is mainly their emotional states and needs extracted from the conversation histories. This is based on our hypothesis that the more similar two partners are, the more likely they can act as mirrors and help each other explore more about himself/herself.
3. In the conversation, partner B first describes his/her own emotions and related events. Partner A can make feedback with either typed ones or the suggested sentences in the suggestion area, based on his/her own considerations. Both participants are encouraged to markup sentences as different emotions with emotion crayons as well as to label B's emotional needs. Thus, as new sentences are typed, the system tries to learn how to make proper feedback according to the situations (events, feelings, needs) and even the feedback guidelines.
4. After a conversation session is finished, both participants will be asked to rate the conversation (1~10). The rating determines how much this conversation contributes to the system's learning process.
   

Architecture & Design Details


Matching people’s emotional profiles
Building the semantic network
What happened?
What emotions may the person have?
What emotions may the person need?
What respond may be good for this person?
Combining the learned feedback with the guidelines?