Clustering news feeds for event groups

  • Dmytro Horobets
Keywords: TEXT, INFORMATION, NEWS, CLUSTERIZATION, CLASSIFICATION

Abstract

The paper considers the processing of information messages. Highlighting news reports, their classification by theme, forming news stories in groups of news, ranking news stories by importance. The proposed algorithm is based on the formation of a set of marker words for messages and a comparison of these sets belonging to different messages among themselves.

References

Data Clustering Contest: Round 1 // site of Developer Challenges / Telegram. URL: https://contest.com/docs/data_clustering (access date: 21.02.2020)

The result of the algorithm’s functioning on test data // site of Developer Challenges / Data Clustering Contest / Telegram. URL: https://entry1178-dcround1.usercontent.dev (access date: 21.02.2020)

Published
2020-03-25
Section
Статті