Nowadays, most businesses rely on the list of trending topics and keywords to decide on major business decisions. These keywords tell them what they need to know – market behavior, current needs or wants of consumers and what’s in and what’s not. It’s faster than undergoing any market research.
To come up with these trending keywords, an algorithm is necessary. Perhaps, the most significant list of trending topics comes from Twitter. Sadly, they have yet to release to the public about the algorithm that they’re using. On the bright side, there’s a simpler algorithm that works in a similar manner. The TF-IDF algorithm, short for term-frequency – inverse document frequency, is also used to find out what the current trending topics are.
TF-IDF has been in use for quite some time. It aids search engines and displays the ones that are frequently used. The more commonly used items ranking higher than the rest.
TF-IDF works with two main concepts – term frequency and the inverse document frequency. Term frequency refers to the number of times a certain term or keyword appears in a single document. Inverse document frequency, on the other hand, works on the notion that the rarer the keyword, the greater its importance in the document. In short, it uses the inverse of the number of documents that contains the keywords. This means that the rarer the word, the greater its weight and vice versa.
In accounting for the list of trending keywords, TF-IDF works this way. It uses a certain target data set against the weights derived from the IDF. TF measures the number of times the target keywords have appeared and the IDF uses the number of documents the keyword appears on. The resulting ranking of keywords will provide the list of trending topics.
The Wonderful World of Algorithms 