Skip to content

Latest commit

 

History

History
65 lines (47 loc) · 5.61 KB

Machine learning for creating web and mobile interfaces on the basis of users' preferences.md

File metadata and controls

65 lines (47 loc) · 5.61 KB
Raw

MACHINE LEARNING FOR CREATING WEB AND MOBILE INTERFACES ON THE BASIS OF USERSʼ PREFERENCES

Date: January 2019

Published here
Link to published document

 

Abstract

The article outlines the author’s method for building a platform that can generate interfaces based on users’ preferences. The algorithms of machine learning that are used and their logical implementation are described. The proposed technology is innovational and actual which will be demanded on the market in the nearest future.

Key words:

  • machine learning
  • UI/UX design
  • web/mobile platforms
  • color theory.

Text

The modern world of technology involves continuous development. A short time has passed since people started using websites and mobile apps every day. One does not have to be explained what a site is. But what do we have every time while creating a regular website or application for your smartphone?

Let's start with a customer. He wants to get in the shortest time possible an Internet solution that works and generates income from advertising. On what does he have to spend large sums of money, starting with the salaries of developers, ending with promotions (someone in this huge web will not even be noticed)? The customer spends insane amounts on a web resource. But what must be done? Of course, there are solutions on the market that provide an opportunity to make an online store in a couple of minutes, but without designers there is no sense to create a web resource. Nobody likes templates.

Let's consider a modern user. He visits web resources and mobile apps to get some information. Information should be presented in the form in which it will be easy to be read and also be adapted for the user and his needs. Those marketers and designers, who know how to do it, earn millions. But how can we predict what the user likes and what doesn't? If there were no modern solutions in IT, customers would have to order complex applications with flexible settings.

The emergence of neural networks and machine learning [1] should change the world. This article describes the author’s unique algorithm that can generate web and mobile applications based on usersʼ preferences.

In fact, the concept of the algorithm is based on usersʼ preferences. And what are usersʼ preferences? These are their favorite sites and applications. We can find out what a user visits, for example, YouTube and Facebook, which is enough to generate one-click web applications for a customer. The customer will only be required to fill in the content, if necessary.

Taking into account the existing algorithms, we propose our algorithm of machine learning:

1. To create a template engine for all main elements on web pages.
2. To get the primary colors for the usersʼ favorite sites.
3. To get each item on the page and compare it with the templates.
4. According to the type of site desired by a customer, for example, a blog, to find all the necessary elements among the analyzed ones.
5. Depending on the usersʼ visits of the same YouTube and Facebook, to collect a page template from the received elements.
6. On the basis on Itten’s color theory [2] to decorate elements, using Aa chromosome algorithm [3].
7. To fill items with content taken out of the databases received from a customer.

As a result, we have an automatically generated solution for the next-generation web applications. The advantages in this case are: we have no designers [4]; no extra money is spent on developers.

Machine learning also allows us, when using a site obtained in the same way, to obtain evolutionary progeny, where no site repeats itself.

Among other things, this solution is optimal for advertisers. Since you can place ad units in the places, where they do not interfere with a user and where he clicks on them more frequently, one can collect more information about the content and the colors, that the user likes.

There is one more aspect. In the case of people with limited abilities this is a whole era of development among web resources. Here is a simple example: a blind person cannot enjoy content, but the proposed algorithm can simply know, that a person does not have the vision to analyze and turn the content into a voice one. People with a poor eyesight will be able to adjust anything on the pages.

This is the future, where advertising can turn into content, where people with disabilities can use any content, where Itten’s color theory will connect Facebook and YouTube colors and give us new colors for the elements on the pages. Isn't it wonderful to live in the world, where everything is developing so fast?

In conclusion, it should be added that the possibilities of this technology are limited only by fantasy and this technology has great potential on the market [5, 6, 7].

Referencies

  1. URL: https://en.wikipedia.org/wiki/Machine_learning (Date of appeal: June 2018).
  2. URL: https://en.wikipedia.org/wiki/Color_theory (Date of appeal: November 2018).
  3. URL: https://en.wikipedia.org/wiki/Genetic_algorithm (Date of appeal: October 2018).
  4. URL: https://css-tricks.com/the-difference-between-responsive-and-adaptive-design/ (Date of appeal: December 2016).
  5. URL: https://uxplanet.org/2019-ui-and-ux-designtrends-92dfa8323225 (Date of appeal: January 2019).
  6. URL: https://www.clock-work.co.uk/blog/general/top-10-most-common-website-problems-and-solutions (Date of appeal: March 2018).
  7. URL: https://www.quora.com/What-is-the-futureof-web-development-for-the-next-5-years (Date of appeal: August 2018).