English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

In this research, we examine the problem of extracting information about domain specific event topics from a large document collection. Given a user query that specifies some seed information about an event topic, a conventional document retrieval system usually finds documents where the given query terms appear frequently but not necessarily the documents containing a complete set of information about the event topic. Documents related to the event topic but not containing the query terms will not be returned. A novel way of representing information about an event topic by its entity and relation instances has been introduced in this thesis. These entity and relation instances may be found in multiple documents. We define an extraction task that involves selecting documents relevant to a given event topic and extracting from them entity and relation instances. This task is known as domain specific event topic entity and relation extraction task. Information extraction patterns are used to extract domain relevant entity and relation instances from documents. We have proposed some document ranking strategies based on these information extraction patterns to address this event topic extraction task. These strategies are known as pattern-based document ranking strategies. Each ranking strategy attempts to assign a score to every document to estimate the gain of event topic related instances found in the document. We have conducted some experiments on three document collection datasets. The three datasets are constructed based on two terrorism events. Experiments have shown that our proposed pattern-based document ranking strategies performed well for document collections of different sizes. A personalized version of the extraction task has been studied. We propose to assign importance weights to entity and relation instances extracted and to personalize the weights based on user feedback. We further extend our document ranking strategies with the personalized entity and relation weighting scheme. We have also developed a working prototype system named INEXT to carry out the domain specific event topic extraction task. 1 ATTENTION: The Singapore Copyright Act applies to the use of this document. Nanyang Technological University Library

Domain specific event topic information extraction on large text collections