Research Areas

Sentiment Analysis

Sentiment Analysis could be defined as one of the mathematical evaluation methods, which give us a Scalar Score of the Distributive Values of an Experience.

For example, we can say that the Sentiment Score of a Image is the Intrinsic Value of the Image determined by the Mathematical Evaluation Function.

Actually, in addition to an image, we can use Sentiment Score for any Modality of Data, without any reference to Externalities.

Thus, Sentiment Score is the Mathematical Measure of Intrinsic Value of the Modality of Data. To put it more simply, it is the measurement of “how much score (value) do you want to give to an object or an experience”?

Sentiment Evaluation is based on the concept that a Tensor T is evaluated in terms of a vector v in Sentiment Space (V3A), which would give us a sentiment score. For more information:

Mathematically, we can define Sentiment Model as a Tensor Evaluation Function (T) with respect to Sentiment Vector Space (v).

The sentiment space is defined as k = 4 Dimensional Vector Space, which is a Linear Space of 4 bases : { Valance, Arousal, Affinity and Accrual }.

We can also find the Sentimentally Similar Images in Sentiment Space.