📜 ⬆️ ⬇️

Emotions, as a criterion for dynamic adjustment of the results of the issue in the search

Does your emotional state change if you cannot find what you need?
I suppose that usually people start to show nervousness and characteristic haste and harshness, their requests change and ...

This idea matured a long time ago and finally took shape when I was shown a quote from Bash, with which this article began:

" single google search history:
s5000vsa fan noise
s5000vsa fans
s5000vsa drown out
s5000vsa fan noise
s5000vsa fun control
s5000vsa silent
s5000vsa how to plug this fucking cart "
reference to the original

Tracking changes in the emotional background of the visitor

Imagine that a search engine tracks the emotional background of a visitor by his behavior on the site. This information is simply added as one of the features that characterize the visitor. Then it will be possible to monitor the changes themselves from the state characteristic of this person and adapt to him.
Adjustment of results

With the results of the issue, which are constantly not suitable for the user, he begins to get nervous, and the system must understand that it is time to give the person other results.

Methods that can be applied for evaluation
What else will it give us

In addition to dynamically adjusting the results, and therefore improving the relevance of the search, this will give the visitor an opportunity to pop up timely messages to the visitor, suitable for his condition.
Generally speaking, when the user is good, it is easier to push him to any great purchases of achievements. And when he feels bad, he can be cheered up. For example, throw links to soothing sites, tint slightly the interface, remind of something pleasant to us. Or show a nice picture.

We understand that by “reading” our mail, search engines basically look for what we are interested in in order to better sell us advertising. But they can use this knowledge and more pleasant.

It can also be used to give up captcha, although I believe that then robots will also learn to "show emotions."

What could this article be called:
Determining the relevance of search results by the user's emotional state.
Evaluation of the relevance of search results by analyzing the visitor's emotions in the search process.
Dynamic assessment of the relevance of search results by analyzing the user's emotions.


Of course, possession of an emotional portrait of a visitor can give us much more than just adjusting the results, but in the end, this is the main thing.
And yes, this analysis can be performed on any site. Imagine that the same Yandex.Metrica shows you not only the gender and age of your visitors, but also their emotions in dynamics when traveling through your site. I think this is very important information, especially for e-commerce.

UPD (11/29/2011) Recognizing the emotional state for telephone services

Source: https://habr.com/ru/post/104483/

All Articles