東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Buckets full of words: Applying auto...

Kersey, Lauren Jenna Elise.

Linked to FindBook

Google Book

Amazon

博客來

Buckets full of words: Applying automatic attribution analysis towards the task of time periodization.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Buckets full of words: Applying automatic attribution analysis towards the task of time periodization./
Author:	Kersey, Lauren Jenna Elise.
Description:	67 p.
Notes:	Source: Masters Abstracts International, Volume: 55-04.
Contained By:	Masters Abstracts International55-04(E).
Subject:	Modern language. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10096540
ISBN:	9781339606484

Buckets full of words: Applying automatic attribution analysis towards the task of time periodization.
Kersey, Lauren Jenna Elise.

Buckets full of words: Applying automatic attribution analysis towards the task of time periodization. - 67 p.

Source: Masters Abstracts International, Volume: 55-04.

Thesis (M.A.)--Saint Louis University, 2015.

The following thesis is a transcript of a three podcast episodes. These podcast episodes summarize the purpose, the plan, and the initial results for an interdisciplinary data science project that I started at Saint Louis University in the fall of 2014. This interdisciplinary project, called the Text Mining Initiative (TMI), includes graduate and undergraduate students of English, Math, and Computer Science. TMI uses data mining, machine learning, natural language processing, and rhetorical theory to explore new approaches to the task of time periodization. Scholars who study historical documents often break up large continuums of time into smaller chunks (time periods). The assumption is often that the properties of these documents (the word forms, for example) change significantly over time yet remain consistent or predictable when scholars observe these smaller chunks. For any large span of time where n equals the number of years, the number of possible time period combinations (meaning the possible combinations of sequential, non-overlapping years) increases exponentially. This project asks whether new technologies can help us select useful categories from these many possibilities. In sum, this project applies the general approach of automatic text classification to experiment with organizing texts into different buckets based on the changes they exhibit over time.

ISBN: 9781339606484Subjects--Topical Terms:

3174390
Modern language.

Buckets full of words: Applying automatic attribution analysis towards the task of time periodization.
LDR:03196nmm a2200301 4500 001 2072911
005 20160902113913.5
008 170521s2015 ||||||||||||||||| ||eng d
020 $a 9781339606484
035 $a (MiAaPQ)AAI10096540
035 $a AAI10096540
040 $a MiAaPQ $c MiAaPQ
100 1 $a Kersey, Lauren Jenna Elise. $3 3188136
245 1 0 $a Buckets full of words: Applying automatic attribution analysis towards the task of time periodization.
300 $a 67 p.
500 $a Source: Masters Abstracts International, Volume: 55-04.
500 $a Includes supplementary digital materials.
500 $a Adviser: Jonathan Sawday.
502 $a Thesis (M.A.)--Saint Louis University, 2015.
520 $a The following thesis is a transcript of a three podcast episodes. These podcast episodes summarize the purpose, the plan, and the initial results for an interdisciplinary data science project that I started at Saint Louis University in the fall of 2014. This interdisciplinary project, called the Text Mining Initiative (TMI), includes graduate and undergraduate students of English, Math, and Computer Science. TMI uses data mining, machine learning, natural language processing, and rhetorical theory to explore new approaches to the task of time periodization. Scholars who study historical documents often break up large continuums of time into smaller chunks (time periods). The assumption is often that the properties of these documents (the word forms, for example) change significantly over time yet remain consistent or predictable when scholars observe these smaller chunks. For any large span of time where n equals the number of years, the number of possible time period combinations (meaning the possible combinations of sequential, non-overlapping years) increases exponentially. This project asks whether new technologies can help us select useful categories from these many possibilities. In sum, this project applies the general approach of automatic text classification to experiment with organizing texts into different buckets based on the changes they exhibit over time.
520 $a This podcast is an additional experiment using audio as a medium for explaining text mining technologies. The first episode presents the research question, outlines the research process, and provides an introduction to the fields of data mining and machine learning. Episode II is a closer look at k-means clustering as one possible approach to periodization, and it introduces the audience to the project's suite of tools. The final episode attempts to interpret these results and concludes by making plans for future steps. Episode I follows my process of learning from the computer scientists on the team; Episode II follows my attempt to share this new knowledge with other English graduate students; and Episode III illustrates how the complexity of these new tools and methods can produce results with their own unique brand of confusion and ambiguity.
590 $a School code: 0193.
650 4 $a Modern language. $3 3174390
650 4 $a Information technology. $3 532993
690 $a 0291
690 $a 0489
710 2 $a Saint Louis University. $b English. $3 1677763
773 0 $t Masters Abstracts International $g 55-04(E).
790 $a 0193
791 $a M.A.
792 $a 2015
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10096540