Headline (Topic generation) with Para
the work is you have to work on a set of plain text and apply the algorithm to find the relevant topics and related to paragraphs
Example: Here is My Sample text
Argentina and Lionel Messi progressed to the last 16 of the World Cup by the skin of their teeth on Tuesday after an 86th-minute strike from defender Marcos Rojo gave them a 2-1 win over Nigeria, eliminating the African side. Croatia advance as winners of Group D with the maximum nine points after beating Iceland 2-1 and Nigeria were just minutes away from joining them before central defender Rojo superbly volleyed home a Gabriel Mercado cross from the right. Messi had put Argentina ahead in the 14th minute, with a fabulously taken goal but the Africans equalized through a Victor Moses penalty in the 51st minute and the twice World Cup winners struggled to respond to that setback with a ragged second half display.
You have to find out like this:-
Topic 1 Argentina strike late to advance to knockout stages
Argentina and Lionel Messi progressed to the last 16 of the World Cup by the skin of their teeth on Tuesday after an 86th-minute strike from defender Marcos Rojo gave them a 2-1 win over Nigeria, eliminating the African side. Croatia advance as winners of Group D with the maximum nine points after beating Iceland 2-1 and Nigeria were just minutes away from joining them before central defender Rojo superbly volleyed home a Gabriel Mercado cross from the right.
Topic 2 Messi Played very well
Messi had put Argentina ahead in the 14th minute, with a fabulously taken goal but the Africans equalized through a Victor Moses penalty in the 51st minute and the twice World Cup winners struggled to respond to that setback with a ragged second half display.
Would love to chat about your project. Let's connect.
Hi, I'm CTO at datascraping [d0t] club would love to discuss your project. I provide whole spectrum of data scrapping solutions, from fully automatic PDF and WEB parsing to advanced Selenium powered + ML powered Computer Vision + captcha solving solutions. Let's chat, I'm the best here! ;-)
What's described in the proposal is less of topic extraction and more of summarisation and more precisely extractive summarisation.
Document summarisation is not an easy problem and their is a large amount of literary work on this topic. There are several existing solutions to the problem which I am confident of delivering if given the opportunity.