This is the submission follow up message
This is the submission follow up message

Perception Metrics Explained

One of the ways that the Rehinged.AI platform interprets content at scale is by measuring “perception metrics” on a corpus (text or groups of text).

A perception metric can be any noun or adjective. The perception metrics are typically selected by the end user to match to the topic and objects of interest. For example, most brand managers have a group of three to seven perception metrics that they want their brand to be known for. A common example for corporate brands is “trust.”

Our AI platform has the ability to interpret and compare any objects on any perception metrics.

Perception Metrics

 

  • Trust

  • Measures the level of trustworthiness. If the text describes uncertainty about the brand, the trust perception score will be lower than for other brands.
  • Enthusiasm

  • Measures the level of enthusiasm or indifference associated with the product. If the text describes the product with passionate words, the enthusiasm perception score will be higher than for other products.
  • Acceptance

  • Measures the level of acceptance or disapproval associated with the product. If the text describes the product with criticisms and objections, the acceptance perception score will be lower than for other products.
  • Value

  • Measures the text mentioned that reflect if the price is worth the features. If the text describes the product as worth its price, the value perception score will be higher than for other products.
  • Uniqueness

  • Measures if the product is viewed as unique or common. If the text describes the product as common and only incremental, the uniqueness perception score will be lower than for other products.
  • Satisfaction

  • Measures the level of satisfaction or dissatisfaction. If the text describes the advantages of the product, the satisfaction perception score will be higher than for other products.
  • Reliable

  • Measures the level of reliability. If the text describes the trustworthiness, durability and quality of the brand, the reliable perception score will be higher than for other brands.
  • Good

  • Measures the level of positivity. If the text describes positively the product, the good perception score will be higher than for other products. This perception metric is most similar to the so-called polarity sentiment score given by common AI algorithms.
  • Like

  • Measures the level of appreciation. If the text describes the endorsement or enjoyment of the brand, the like perception score will be higher than for other brands.
  • Desire

  • Measures the level of need. If the text describes the craving of the brand, the desire perception score will be higher than for other brands.
  • Durable

  • Measure the level of durability or how enduring something is. If the text describes the brand as ephemeral, the durable perception score will be lower than for other brands.
  • Natural

  • Measures the level of pureness. If the text describes the alterations of the product, the natural perception score will be lower than for other products.
  • Ethical

  • Measures the level of ethics. If the text describes the ethics association with the brand, the ethical perception score will be higher than for other brands.

 

 

Perception Metric Score Disclaimer

Our AI platform includes a renormalization of the perception metric score depending on the length of the text. In other words, the perception scores between two similar texts of very different lengths will be similar.

The perception score reflects the positive or negative point of view of the text with respect to a perception metric, relatively independent of its length. However, longer texts have naturally the tendency to describe pros and cons (multiple point of views) and not only one point of view, so the perception score might be more neutral for longer text than shorter ones.

For example, a major brand is discussed in 10x more articles than a smaller brand and in various contexts. The thing to keep in mind is that the overall score for a more discussed topic will statistically start to average more toward neutral.