Please use this identifier to cite or link to this item:
Title: Persons of interest
Authors: Grech, Nigel
Keywords: Information retrieval
JavaScript (Computer program language)
Text processing (Computer science)
Issue Date: 2016
Abstract: In our culture there is a deep interest in the lives of public figures and noteworthy individuals. There are genres of books, websites even television and YouTube channels dedicated to examining their lives. Timelines are a visual medium through which the moments in these individuals' lives can be expressed. However for the researcher, journalist or educator who has the task of compiling these timelines a great deal of time and e ort is required. To our knowledge there exists no automated system that can analyse text documents to generate a human readable timeline. In this research, we propose a solution to this problem by providing a system which analyses a source document and generates a timeline of all the relevant moments in a targeted person's life. The method proposed consists of three main components: firstly obtaining and managing relevant structured information. Secondly obtaining unstructured text and preprocessing it by means of tokenization, part of speech tagging and named entity recognition. After collecting and preparing the required data, all relevant information needs to be extracted from it. This final step will involve extracting dates and related phrases. The relevance of these phrases to the timeline will then be determined, this process will be aided by comparing phrases with statements obtained from structured information. In cases where structured information provides no assistance the system will then have to determine if a phrase is relevant to the individual in question and the context of a timeline. Finally the events that have been extracted need to be compiled into a timeline and visualised for a human reader. We do this through the use of a JavaScript application that allows the user to interact with the timeline through a browser. In this prototype we have mainly focused on politicians and public figures. Working with data obtained from Wikipedia and WikiData. However such a system can be extended to work in different text sources and different types of people.
Description: B.SC.IT(HONS)
Appears in Collections:Dissertations - FacICT - 2016
Dissertations - FacICTAI - 2016

Files in This Item:
File Description SizeFormat 
  Restricted Access
708.35 kBAdobe PDFView/Open Request a copy

Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.