Tihomir Čuljak
Internet browsers or search engines are particular websites that can help to locate specific information from other sites. Information can mean web pages, images, movies, files, databases, social networks, open-source directories, etc.
Web directories are maintained by people as text editors, while Internet browsers work on a specific mathematical algorithm that can be combined with human intervention.
The first tool used to search the Internet created in 1990 is known as Archie. It searched for file names and headlines but did not index the contents of those files. The first online search engine that indexed the content was created in 1993 and was called Wandex.
One of the first internet browsers of the full text was made in 1994 and was called WebCrawler. WebCrawler has allowed users to search any word from any site. Soon after that, many search engines like Excite, Infoseek, Inktomi, Light, AltaVista, Yahoo, MSN Search, and Google appeared.
Google has become the most popular search engine around the 2000s and can do searches through keywords linked to the operator AND, OR and other less popular operators.
Google has developed a PageRank tool that arrays websites based on the number and PageRank of other sites, including sites that link to them, assuming that more popular websites also link better websites.
Besides, as some page was more popular and more visited, it was more at the top when the user wanted to find a specific term on that page.
Internet browsers collect different information about websites. They include websites in their database, and offer to any user who is looking for a particular term, specific service, product or something else.
The Internet Explorer consists of following: a data collection program from other websites, an index containing the terms collected by the collection program during a search of a website and an index search engine.
The usefulness of the search engine depends on the relevance of the results it provides.
Although there may be millions of websites that include a specific word or phrase, some pages may be more relevant, popular or more reliable than others.
Most search engines apply methods of displaying or ranking the results in order. By listing the results in order, their goal is to obtain the "best" results. Certainly, some search engines differ from other search engines.
Wolfram Alpha is a knowledge search engine that indexes a lot of different criteria that cannot be combined.
Meta-browsers are web browsers that simultaneously use more search engines at once and show the results of each search engine.
Some websites that are used to put out certain mass content stored in their databases, similarly search their databases.
The most known such web address is YouTube that collects movies from users and allows search by other criteria and not just by keywords.
Considering that search in all mentioned search engines is most commonly performed via keywords, and less frequently by subject, they all represent linguistic search engines.
The problem with linguistic search engines is that they search for content in one language.
In addition to searching by keywords (tags), algorithms measure the intensity and priority of pages where keywords are found. That is why possible manipulations with results can occur, whether they are installed by algorithms or organized by a competitor who wants to hide specific content of market opponents.
By using algorithms, some content and pages can have a higher priority, so that they appear at the top of search results.
Those who pay for advertising and sites that are more frequently opened, especially in the near past usually have a higher priority.
In a case when market competitors want to hide some specific site, they can do so by publishing content on websites daily. That daily published content usually has the same keywords in the title, at the beginning of the article or in tags as the content they aim to hide. Thanks to this, a user who wants to find a page he or she has seen one or more days ago, cannot do this unless he has saved a specific link to the page or knows the exact name of that page. Therefore, this represents a kind of censorship of content that somebody wants to hide.
The essential weakness of all today's search engines is that words they see as a line of letters and do not recognize the meanings of particular words.
As artificial intelligence does not know the meaning of the word, it cannot even know the meanings of phrases or sentences.
If we write a paragraph in a translation program, the translator will translate it very well into another language. However, if this translation is translated back into the original language, we will get a significantly less meaningful text. If we repeat this 6-7 times, we will get a text that does has no meaning.
The meanings of words can be either sensory, emotional or abstract. The most critical sensory meanings are visual or optical, acoustical meanings, scents, tastes, and haptic meanings. Also, there are a lot more emotional meanings. Emotional meanings describe the emotions that an individual word provokes in the reader or listener.
The visual meanings of individual words can be defined with images or short motion movies.
The acoustical meanings of particular words can be defined as the pronunciation of words.
During the reading of the specific word, that word can be associated with acoustical, visual and other alternative meanings.
On the other hand, when we hear a certain word, we can also find information about whether the person who pronounced that word is someone we know. Additionally, we can tell if that person is an adult, a woman, a child, etc. It is possible to extract more and different information in the same way from other types of word meaning.
The meanings of each word vary by priority, intensity and time order. Some specific meaning in the sentence might be more critical and more important from the other meaning.
Also, in some other given sentence, a fifth meaning may have the highest priority, which depends on the context of the word in a sentence or a paragraph of the text.
The intensity level of the word meaning can alter the priority of a particular word in a sentence or sentence in a longer text. Therefore, the intensity level can provide information regarding the importance of words in a specific context. Moreover, words in a string can also represent different time frames of events, by which the image, sound or other meaning changes over time.
For artificial intelligence to correctly translate databases that are used for translation, apart from the words in different languages, it would have to create a tabular representation of all possible meanings for each word.
Next to each letter, word, phrase or expression, there should be an image of that word or phrase in several variants, short film, sound, the label of the taste and smell, and all other meanings that a letter, word, phrase or expression may have.
Only then could the artificial intelligence be able to translate the words into a "meaningful spatial-temporal four-dimensional film" that would re-translate those words into the other language.
To overcome the shortcomings of keyword-based content search that artificial intelligence does not understand, I have created a platform for users to publish their articles or search for other articles by some specific important meanings, no matter what the language is.
Search by keywords often comes with thousands of solutions. Therefore, to better filter the results, I have imagined that content authors who want to reach the target readers more easily, before publishing their article, could have their texts tagged with several different meanings (criteria - tags) that can be combined.
The tags are connected with the AND operator and by choosing more and more criteria, the number of filter results decreases whilst the precision of the search increases.
In the first attempt at the beginning of 2009, I made a website where I have enabled potential authors to label their pages according to approximately sixty criteria divided into four drop-down menus (theme, style, for whom, page type) so that users from each drop-down menu can choose one tag or criteria. The selected criteria were connected with the operator AND.
The site was imagined as a sort of web directory.
Users would use the filter to search for texts in which authors indicated links that describe in more detail what the users search for by logical search, and according to entered tags. This attempt has failed, as the site was designed not as a social network, but rather only for the search of entered tags and then referring users to other sites. Hence, not many people were interested in it.
At the beginning of 2018, I decided to try again and engage developers who would make a web platform according to my instructions. This new attempt is imagined as a social network where every registered user can put their texts as a medium or tool for their own promotion, promoting their ideas, or commenting on other content.
Each site visitor can track other authors, create the following list and add wanted site authors on it or can search published articles by 134 available criteria or tags.
Selected criteria can be combined with a simultaneous selection of multiple search criteria. With this kind of search, the precision can increase a million times, and the search results are the most accurate. How precise the search will be, depends on how accurately the author tags or labels his article, but also on who is searching for the wanted content.
The site is registered under the domain named Novenews.net and is imagined as a combination of news portals, blogs, social networks, forums, and Facebook, with the ability for every registered user to publish their articles.
Anonymous visitors can search for the entered text, i.e. article by selecting shortcuts on the home page, clicking on the wanted author, or by choosing filter criteria.
By selecting the author you can get all of his/her texts that have been published. Articles and comments are displayed by the date of publication. The purpose of this kind of categorization is to encourage discussions similar to those on different forums.
Registered users can publish their own articles, comments on other papers, and choose the authors whose publications they want to follow.
Authors that register users decide to follow are extremely easy to find in their drop-down menu, within a list of all followed authors.
The registered users can search for all articles in the same way as anonymous users.
When writing a new text that a registered user wants to publish, he or she links the tags that best define his / her text. Tags are divided into 6 groups and in each group, one or more criteria (tags) can be selected.
The TARGET GROUP criterion contains tags to indicate who we intended our article for.
The OCCUPATION criterion indicates the occupation or profession of the targeted readers.
The THEME criterion contains topics that make the article different from other available topics.
The STYLE criterion contains the style of the written text.
The EMOTION criterion contains a list of emotions that the text expresses or encourages.
The SENSE criterion contains a list of senses that text mentions, describes, or supports.
When a writer chooses all the tags, he uploads the text to the web page.
The writer can choose up to 4 criteria in each menu, or 4 in each group of labels.
After the text writer chooses all the criteria that he wants to tag his text, the text is uploaded and published in the database.
There are several ways to search for articles in the database.
The first way is by using shortcuts: News, Sports, Tech, Science, and Culture.
By selecting wanted shortcuts, you get articles from several topics that can be categorized into these major groups, which readers are most interested in. Published texts are sorted by the date of entry and the number of readers that have read the text.
Another way of searching is by the author. The reader can see all the publications of the selected author by merely clicking on the author's name. The desired article can also be opened by clicking on it.
The articles are listed by the date of the entry and the last comment.
If the reader wants to keep track of what the author he is interested in is writing in the future, he/she can click the button subscribe or follow. The author's name then is placed in the list of authors that this registered user follows. In that way, it is easier to follow articles that the user is interested in.
By clicking on the author's name, all the articles of that author are dispayed by the date of the entry and the last comment. The purpose of this kind of categorization is to encourage discussion as in the forum.
The third way to search for articles is by the filter. In the filter, the user opens each group of tags, and in each group chooses available criteria. By selecting the desired criteria and then clicking the SEARCH button, all the articles that match mention criteria are displayed. If the user selects a large number of available criteria, the number of articles decreases, whilst the precision of the search increases.
The advantage of this search engine is that the publications that the reader searches for can be obtained from the database, regardless of the language they are written in and when they are written. Since the results are a combination of 6 different criteria groups, the precision of the search is up to one million times better than when the search is conducted by only one criteria group that cannot be combined. If the reader gets results in a foreign language, they can easily be translated by a translator, preferably Google Translator. If a user selects one criterion in a filter, he /she will get a list of all the articles that are tagged, among other tags, with that tag.
If the user chooses that, but also one other criterion, a slightly smaller list of articles that are marked with both selected tags will appear. If a third filter criterion is selected, even a smaller list of articles with all three tags will appear.
By increasing the number of selected filter criteria, the precision of the search is also increased. All selected criteria are linked to operator AND, which selects articles with all selected tags.
That allows the user to accurately determine which topic he/she is interested in, what type of reader the text targets, what is the style of the writing, what level of reader's expertise the text is written for, which emotions the text describes, mentions or triggers, and which senses text mentioned, describes or stimulates.
In the future, this logical filter could add another set of criteria that artificial intelligence can evaluate, such as, for example, the date of the article publication, etc.
This would eliminate the weaknesses of various newspaper portals where the news lasts only for a day, after which is quite challenging to find that same news in the near future. Also, social networking weaknesses would be surpassed. Those weaknesses are that users usually receive information about the last article or publication, image or movie recording of the person he/she follows, while earlier publications require additional manual search without knowing exactly what is the content is.
Other types of web portals could include criteria such as electronic types of texts, movies or music content, like YouTube, and location or language filters that can be recognized by artificial intelligence, etc. This could eliminate the weakness of existing search engines that cannot find content that written in another language or something that is hidden by millions of other more recent publication that have the same titles and tags.
How to become a free, independent, occasional and hobby journalist
If you want to publish a particular article on a specific website, you can do it only by commenting on some other publication.
Furthermore, if you publish your article on social networks such as Facebook, Twitter, etc. your text will read those who follow you, or in a case that someone shares your article, possibility that a larger group will read is greater.
If you want to forward the content just to those who are interested in a specific topic, this can be done on the new site Novenews.net/hr.
This page looks like a web portal. However, unlike the usual web portal, everyone here can be a journalist. To write articles, it is merely enough to be registered. Registration is very simple.
Click on the ADD POST (news) button and the form to sign in or register will appear. You must enter a name or a nickname, your email address for receiving a registration email that you will have to confirm. Your email may end up in Spam, so if you do not get it, check your Spam folder. This is most likely to happen with emails with the extension @gmail.com.
Once you have selected your article for publication on the web portal, you can tag it with 132 criteria that allow readers to find exactly what you have been writing.
The user can select and publish the profile picture with each of their texts. However, that is not obligatory. In order for your articles to be found by the very readers you targeted, it is important to have the most accurate selection of tags. The tags are divided into 6 groups and in each group you can select one, none or more criteria (tags).
The TARGET GROUP criterion contains tags to indicate who we intended our article for.
The OCCUPATION criterion indicates the occupation or profession of the targeted readers.
The THEME criterion contains topics that make the article different from other available topics.
The STYLE criterion contains the style of the written text.
The EMOTION criterion contains a list of emotions that the text expresses or encourages.
The SENSE criterion contains a list of senses that text mentions, describes, or supports.
When a writer chooses all the tags, he uploads the text to the web page.
When the article is uploaded, readers can find it for a shorter time on the home page. When the other more recent articles are published, readers can find it by clicking on the Filter, where the reader chooses tags of his choice.
Selected tags within one group can be combined. Criteria within a specific group are linked with the operator OR, meaning that it's enough to select only one tag to find your article.
Selected tags from other groups are linked to selected tags from previous groups with operator AND, which means that your article will appear in the results only if you have selected the right tag in all the groups.
By choosing more tags from different groups, the number of filtered articles decreases, whilst the precision of the search increases.
In addition, articles can be filtered by 6 major groups or by clicking on the author's name. Authors whose content you want to follow regularly, you can select ''follow me'' option. Additionally, if you have written some texts, you can see how many users follow your content. Publications can also be searched by keywords in the title or author's name, and the form for this search is at the top of the filter. By clicking on the author's name, all the content of that author appears, displayed by the date of the publication and the last comment.
This allows the author to see which article was last commented so that he/she could comment it back.
The advantage of this search engine is that the content that the reader searches for can be obtained from the database of publications, regardless of which language they are written in and when they are written. Since the results are a combination of 6 different criteria groups, the precision of the search is up to one million times better than when the search is conducted by only one criteria group that cannot be combined.
If readers are interested in news that is current for only one day, they might find it here, but if readers are interested in stories and articles that are trending for a longer time, it would be better to visit this site more often.
Tags
Featured articles
