Affective Adjustment Materials Selection Based On Ordered Weighted Averaging Operator

According to affective adjustment process theory which is proposed by Gross, affective adjustment process mainly includes five steps, they are situation selection, situation modification, attentional deployment, cognitive reappraisal and expression suppression. Reappraisal and suppression are both popular regulation strategies. Moreover some researches have shown that when compared with expression suppress, cognitive reappraisal is more effectively to regulate people’s negative affections. Cognitive reappraisal is that people modify an existing and negative situation so as to obtain a different level of affection such as positive affection by watching TV, playing games and listening music etc. However, there must be a lot of adjustment materials to support it. What’s more, it will consume a lot of time to find and select the materials. So it is inconvenient for some people especially for computer users who need to work effectively long-term, and don’t have enough time to find the materials. So it is very important to establish a mechanism which can select a suitable material to regulate people’s negative affections based on characteristics of material and the people’s preferences. This paper presents a computer-mediated affective adjustment method, which mainly includes a material evaluation model and material selection model. Besides, the model will be evaluated from both subjective and objective aspects. Based on users’evaluation, a multi-attribute decision algorithm will be used to select an appropriate adjustment material that will be used to help users to regulate negative affections.So my main work as follows:1. Establish an affective adjustment material library. According to the characteristics of affection regulation which based on human-computer interaction,35materials are selected from four aspects to create an affective regulation material library. And the four aspects respectively are affection valence, material type, length of play time and the presentation. 2. Establish evaluation matrix of material attributes. According to material evaluation method in psychology and the characteristics physiological signal, a four-dimensional material attribute set is founded. Besides, an evaluation matrix of35materials is also to be constructed.3. Set up a material selection model based on ordered weighted averaging operator (OWA). Propose to use a kind of multiple attribute selection algorithm-ordered weighted averaging operator, to complete the material selection process. And introduce the concept and characteristics of OWA operator specifically. Finally construct a model based on the OWA operator, and put forward the problem which the model needs to solve is how to weight for material attributes.4. Weight for materials’attribute and create the weight vectors. According to the characteristics of material attributes and the in-depth study of subjective weighting and objective weighting method. A combination weighting method based on the sum of squared deviations and the method of interactive weighting based on BP neural network are used to weighting for materials. In the end four weight vectors of35materials based on the four kinds weighting methods are constructed.5. Verify the validity of OWA operator. According to the evaluation matrix and weight vectors. OWA operator was to be used to make the final decision to select an optimal material. On the one hand we make a simulation experiment on Matlab to make sure the results of the ordered weighted average operator which based on the above4kind weighting methods can effectively eliminate the five materials with no moderating effection, and the18th material is selected to be the best regulation material. So that it is verified that the OWA operator is effective on selecting affective adjustment material. On the other hand an experiment was designed to verify the method of interactive weighting based on BP neural network was better than the method of combination weighting based on the sum of squared deviations.6. Construct an affective regulation system with the function of affective support according to the affection recognition based on physiological signals and the material selection model.