PROJECT TITLE :

Sequential Summarization: A Full View of Twitter Trending Topics

ABSTRACT:

Every day, Twitter collects millions of tweets as an information delivery tool. However, when confronted with the overwhelming and disorganized messages, some users, particularly new users, find it difficult to understand hot topics on Twitter. Existing work has attempted to provide a small excerpt to explain a topic, but this only provides limited benefits and fails to meet the expectations of consumers. In this research, we propose a novel summarizing job, sequential summarization, which attempts to produce a series of chronologically ordered brief sub-summaries for a trending topic in order to provide a complete story about the topic's evolution while preserving the information presentation order. The amount of sub-summaries for distinct topics is not predetermined, unlike the standard summarizing assignment. To determine the essential subtopics inside a trending subject, two approaches are developed: stream-based and semantic-based approaches. Then, for each subtopic, a brief sub-summary is prepared. In addition, we offer three new measures for evaluating the system-generated summaries' position-aware coverage, sequential originality, and sequence correlation. The effectiveness of the offered methodologies has been proved by experimental findings based on the given evaluation criteria.