Automatic Information Extraction about Persons from News Articles

The current paper is devoted to the autiomatic extraction of information about persons from news articles Automatic analysis of texts and writing genres has been of particular interest ever since the appearance of the computers. Nowadays Natural Language Processing (NLP) is widely used in a number of domains with the aim to computerize human language studies. For instance, one of the most important applications of NLP is the NER (Named-entity Recognition) which tools seek to locate and to classify named entities in text into pre-defined categories. The aim of the current study is to develop a general-purpose algorithm for the automatic analysis of newspaper articles based on the social graphs theory. The study solves several tasks, such as collecting newspaper articles from a website, identifying characters in a newspaper article, representing the relationships between characters, creating the desktop user interface. The result of the present study is a desktop application, permitting to perform the analysis of news articles (texts) focused on the relations between characters. The program automatically extracts characters’ names from the text, associating them with their clusters. Basing on the data, it is possible to build a graph representing the relationships between the characters and determine its main characteristics. It is important to note that the current product might be used in any domains related to the analysis of the text and the definition of personal names.