Show simple item record

Files in this item

Cover for Topic Flow Model: A Graph Theoretic Temporal Topic Model for Noisy Mediums
dc.contributor.advisorSingh, Lisa Oen
dc.creatoren
dc.date.accessioned2017-09-18T15:39:43Zen
dc.date.available2017-09-18T15:39:43Zen
dc.date.created2017en
dc.date.issueden
dc.date.submitted01/01/2017en
dc.identifier.otherAPT-BAG: georgetown.edu.10822_1044619.tar;APT-ETAG: 3ec093fdb7849c0cd7f9e5933d7ad021; APT-DATE: 2017-10-25_10:53:04en-US
dc.identifier.urien
dc.descriptionM.S.en
dc.description.abstractIn the modern era, data is being created faster than ever. Social media, in par-en
dc.description.abstractticular, churns out hundreds of millions of short documents a day. It would be usefulen
dc.description.abstractto understand the underlying topics being discussed on popular channels of socialen
dc.description.abstractmedia, and how those discussions evolve over time. There exist state of the art topicen
dc.description.abstractmodels that accurately classify texts large and small, but few attempt to follow topicsen
dc.description.abstractthrough time, and many are adversely affected by the large amount of noise in socialen
dc.description.abstractmedia documents. We propose Topic Flow Model (TFM), a graph theoretic temporalen
dc.description.abstracttopic model that identifies topics as they emerge, and tracks them through time asen
dc.description.abstractthey persist, diminish, and re-emerge. TFM identifies topic words by capturing theen
dc.description.abstractchanging relationship strength of words over time, and offers solutions for dealing withen
dc.description.abstractflood words, i.e., domain specific words that pollute topics. We conduct an extensiveen
dc.description.abstractempirical analysis of TFM on Twitter data, newspaper articles, and synthetic dataen
dc.description.abstractand find that the topic accuracy and signal to noise ratio are better than state of theen
dc.description.abstractart methods.en
dc.formatPDFen
dc.format.extent97 leavesen
dc.languageenen
dc.publisherGeorgetown Universityen
dc.sourceGeorgetown University-Graduate School of Arts & Sciencesen
dc.sourceComputer Scienceen
dc.subjectData Miningen
dc.subjectGraph Miningen
dc.subjectMachine Learningen
dc.subjectNatural Language Processingen
dc.subjectText Miningen
dc.subjectTopic Modelingen
dc.subject.lcshComputer scienceen
dc.subject.otherComputer scienceen
dc.titleTopic Flow Model: A Graph Theoretic Temporal Topic Model for Noisy Mediumsen
dc.typethesisen


This item appears in the following Collection(s)

Show simple item record