Abstract— This paper tries to explain the concept of hash tags and how the trending hash tags are determined by Twitter. To explain the working we have tried to analyze and algorithm from a source and on the basis of that we are trying to understand the steps taken in computing the trending topics on twitter. We are also presenting with some methods for users to find the trending topics on twitter.
Keywords—hastags; Twitter; trending;
Introduction
It’s the age of Internet. Twitter is the new singing birdie. Twitter allows us to know not only what’s hot in our area but in the world. But how do we know it. Or rather, how Twitter knows the shift of tweets. So let’s try to understand the method employed by Twitter to compute the trends of tweets and also how we can find what’s trending
Trending tweets: How twitter finds them
Hashtags
Finding the trend
Twitter has not publically made its algorithms available. But to find the trend, following general steps need to be followed:
First classify the tweets on the basis of location.
Now take the hash tags from the tweets of a particular location and classify them into various topics based on the algorithms like Latent Dirchlet Allocation Algorithm.
Find the popular topics in a location and the most popular hash tag of that topic.
It is also necessary to find correlated hash tags.
Finally, the trending hash tags are those which are
Popular in the tweets of the same topic
Belong to the most popular topics
Finding Trends: How we can do it
Twitter saves us from all this overhead explained above. It already presents a list of trending topics in a location (Ten trending hash tags). It “tailors” the trending hash tags according to our location but we can change the location and find what’s hot there.
Besides this, there are some tools or software that help us with the same like Bundlepost, Hashtracking etc.
References
[1] M. Mathioudakis and N. Koudas, “TwitterMonitor: Trend Detection over the Twitter Stream,” in SIGMOD:International Conference on Management of Data, New York, NY, USA: ACM, 2010, pp. 1155-1158”
[2] Twitter FAQs https://support.twitter.com/articles/101125
