Track #hash and create description from it [LOGICA]

Asked

Viewed 161 times

0

I have come across this now and I would like to debate to make a logic.

Well, nowadays it is used a lot the hashtags for everything in the main social networks, I almost never enter in Twitter and when I come across strange hashtags and do not know what they are talking about, it leaves the user a little lost, what if there was a mechanism that created a description of such a hashtag? To describe in small lines what they are saying to show the user and not leave lost.

I remember reading a news that Twitter was planning/releasing something like this but I could not find more information, the "mechanism" would have to follow 3 rules of practice:

  • Be fully automatic, without human help to create the description
  • Be programmed in PHP and Mysql
  • Maximum 200 characters for description.

With my thoughts off I imagined that we would have to take the messages that contain the HASHTAG X in the database, see words that have in common, from this he searches in a database and assembles the description, I think so and you?

php mysql sql twitter

  • I did not understand why the negative without a reason. ANYWAY.

  • The site is for questions and answers lenses on issues techniques. This question is broad, discursive and outside the scope of programming. It does not fit in our focus: http://answall.com/help/on-topic. I am voting to close as "primarily based on opinions". But aside from the topic, there is nothing wrong with the question, there is no reason to deny.

  • @Guilhermebernal would be a question with a useful answer that would help many people in the future since my focus is how to do this in Mysql and PHP, would not be interesting?

  • Still, it is always necessary to be able to choose an answer as "The correct answer". Moreover the question must be clearly objective and without any kind of personal opinion. Opinionated questions do not produce answers, they only produce a lot of opinions that are not one better than the other. If the focus is how to do it in Mysql and PHP, then the question does not demonstrate this focus. And yet, it would be too broad because you ask how to do the whole system, there’s no way to answer that. A "correct and definitive" answer would be a book, or would make the system for you. Understand?

  • Anyway, you can open a question there at the [goal] which is where we discuss the community itself. There would be the place to ask about this question and discuss whether or not it should be accepted. All our rules can be changed/improved by discussion there.

1 answer

1


Assuming the question is actually on-topic and not based on opinions (I’m not entirely sure, but I’m giving you the benefit of the doubt), it seems to me that what you want is to implement in PHP and Mysql an algorithm that invents automatically, for each hashtag a description of no more than 200 characters for it. That’s it?

If you are right, your problem falls into a very complicated area of artificial intelligence, as the system would have to analyze the texts related to each hashtag, understand what they mean and what their relationship is with each other and then come up with a description that makes sense.

That is, your algorithm will have to be able to understand a text where someone celebrates the goal, or complains of a certain politician or speaks what he thought of the novel, or tells about the trip abroad, or tells an internal joke of a certain group... To make matters worse, people often write texts with poor spelling, write in several different languages and use hashtags in a totally incoherent way. I mean, you better forget this idea.

With the algorithm you’re guessing, the most you can get is a bunch of words that you know appear in the texts with the hashtag. But, so what? A heap of words of unknown meaning alone does not form an intelligible description of anything.

  • it would be quite complicated, but I followed a less artificial logic and managed to create the algorithm that makes it, thank you very much!

Browser other questions tagged php mysql sql twitter

You are not signed in. Login or sign up in order to post.