A Study on Timeframe based Radical Event Detection on Online News Archive
Journal Title: Engineering and Technology Journal - Year 2020, Vol 5, Issue 7
Abstract
In the field of Natural Language Processing, Event detection has been an active area of research. While most of the work emphasizes on detecting all possible events by using datasets comprising of news article belonging from a broad range of time and location (collected either from a single data source or multiple data sources), this work focuses on detecting an event using short timeframes of News Articles extracted from an Online News Archive. Also, this is a non-targeting approach when it comes to detecting the theme or category of event it focuses on but becomes a targeted one as it tries to put all its focus on detecting one major event when applied on a timeframe. While other approaches are based on various techniques like semantic graphs, clustering algorithms, topic detection and tracking for event detection, this approach detects events with the help of a filtering mechanism combined with a novel threshold. This algorithm leverages detection of trending characteristics of an event in a timeframe and filtering news articles which show similar characteristics. And finally, we evaluate our algorithm with precision, recall and the percentage of articles which are actually related to the event in a timeframe termed as dominance percentage.
Authors and Affiliations
Pritish Gupta , Prof. Sridhar Ranganathan
Keywords
Related Articles
- EP ID EP698299
- DOI 10.47191/etj/v5i7.02
- Views 94
- Downloads 0
Impact Load Resistance of Steel after Heat Treatment and Quenching with Distilled Water
Given that low carbon steels have favourable mechanical properties, they are widely used in power plant components, oil and gas pipelines, and armor structures. However, the relationship between a steel's structure and m...
INVENTORY COST ANALYSIS WORK IN PROCESS (WIP) UNDER FOCUS OF LEAN ACCOUNTING, SIMULATION AND OPTIMIZATION SCENARIOS
In this research, we present a cost analysis of the materials that are in a queue to be processed, better known as work-in-process inventory (WIP, for its acronym in English).The inventory in process will be analyzed wit...
DESIGNING FRAMEWORKS FOR RELIABILITY IN DEEP LEARNING SYSTEMS
There has been a great amount of progress in deep learning models in the last decade. Such models are most accurate when applied to test data drawn from the same distribution as their training set. However, in practice,...
Analysis of the Level of Satisfaction and Importance of Contractor Performance Factors on Construction Projects
In the midst of increasingly tight competition in the construction world, every contractor company is required to be able to demonstrate the quality of its services so that it can excel in the competition. As a company e...
EVALUATION OF THE MEDIUM VOLTAGE CABLE LINE (SKTM) 20 KV USING TAN DELTA AND PARTIAL DISCHARGE METHODS
Distribution network with cable system requires maintenance to maintain the level of reliability. Many factors affect the reliability of this cable system, one of which is environmental factors that can cause deformation...