서브메뉴
검색
Time-Series Analysis of Sentiment in Social Media Can Predict Individual and Collective Behavior of Public Health Significance- [electronic resource]
Time-Series Analysis of Sentiment in Social Media Can Predict Individual and Collective Behavior of Public Health Significance- [electronic resource]
상세정보
- 자료유형
- 학위논문파일 국외
- 최종처리일시
- 20240214101549
- ISBN
- 9798380138796
- DDC
- 614
- 서명/저자
- Time-Series Analysis of Sentiment in Social Media Can Predict Individual and Collective Behavior of Public Health Significance - [electronic resource]
- 발행사항
- [S.l.]: : Indiana University., 2023
- 발행사항
- Ann Arbor : : ProQuest Dissertations & Theses,, 2023
- 형태사항
- 1 online resource(298 p.)
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-02, Section: B.
- 주기사항
- Advisor: Bollen, Johan;Rocha, Luis M.
- 학위논문주기
- Thesis (Ph.D.)--Indiana University, 2023.
- 사용제한주기
- This item must not be sold to any third party vendors.
- 초록/해제
- 요약Lexical sentiment analysis has been used to understand what is expressed in natural language, sometimes to better understand the psychological characteristics of an author, or to understand the author's stance towards an object, such as a product, person, or idea. This methodology has also been used to study temporal patterns in the mood of populations, or to understand the general use of language. In general, these applications study changes in the central tendency of the mood of entire populations. However, it is likely that collective moods may be composed of discordant parts. Can we use these sentiment tools to predict and understand health outcomes for both small cohorts and at the level of populations, and does the analysis of the distribution of sentiment reveal distinct components useful for those goals?In this dissertation, I demonstrate how the use of natural language processing and lexical sentiment of social media timelines can be useful in predicting health outcomes for a small cohort of epilepsy patients. I develop a method based on the singular vector decomposition to discern characteristic components of the distributions of collective sentiment, associated with sub-populations and cohorts of interest. I demonstrate that the first singular component represents the base distribution of sentiment due to the frequency of sentimental words in natural language and show how further components can reveal meaningful patterns in sentiment over time. To show the predictive and analytical power of these components, I demonstrate their use in modeling sex searches as a proxy for human reproductive cycles and mortality during the Covid-19 pandemic.
- 일반주제명
- Public health.
- 일반주제명
- Computer science.
- 일반주제명
- Information science.
- 키워드
- Complex systems
- 키워드
- Social media
- 키워드
- Natural language
- 기타저자
- Indiana University Informatics
- 기본자료저록
- Dissertations Abstracts International. 85-02B.
- 기본자료저록
- Dissertation Abstract International
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008240612s2023 us |||||||||||||||c||eng d■001000016934270
■00520240214101549
■006m o d
■007cr#unu||||||||
■020 ▼a9798380138796
■035 ▼a(MiAaPQ)AAI30574110
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a614
■1001 ▼aWood, Ian Benjamin.
■24510▼aTime-Series Analysis of Sentiment in Social Media Can Predict Individual and Collective Behavior of Public Health Significance▼h[electronic resource]
■260 ▼a[S.l.]:▼bIndiana University. ▼c2023
■260 1▼aAnn Arbor :▼bProQuest Dissertations & Theses, ▼c2023
■300 ▼a1 online resource(298 p.)
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-02, Section: B.
■500 ▼aAdvisor: Bollen, Johan;Rocha, Luis M.
■5021 ▼aThesis (Ph.D.)--Indiana University, 2023.
■506 ▼aThis item must not be sold to any third party vendors.
■520 ▼aLexical sentiment analysis has been used to understand what is expressed in natural language, sometimes to better understand the psychological characteristics of an author, or to understand the author's stance towards an object, such as a product, person, or idea. This methodology has also been used to study temporal patterns in the mood of populations, or to understand the general use of language. In general, these applications study changes in the central tendency of the mood of entire populations. However, it is likely that collective moods may be composed of discordant parts. Can we use these sentiment tools to predict and understand health outcomes for both small cohorts and at the level of populations, and does the analysis of the distribution of sentiment reveal distinct components useful for those goals?In this dissertation, I demonstrate how the use of natural language processing and lexical sentiment of social media timelines can be useful in predicting health outcomes for a small cohort of epilepsy patients. I develop a method based on the singular vector decomposition to discern characteristic components of the distributions of collective sentiment, associated with sub-populations and cohorts of interest. I demonstrate that the first singular component represents the base distribution of sentiment due to the frequency of sentimental words in natural language and show how further components can reveal meaningful patterns in sentiment over time. To show the predictive and analytical power of these components, I demonstrate their use in modeling sex searches as a proxy for human reproductive cycles and mortality during the Covid-19 pandemic.
■590 ▼aSchool code: 0093.
■650 4▼aPublic health.
■650 4▼aComputer science.
■650 4▼aInformation science.
■653 ▼aComplex systems
■653 ▼aSentiment analysis
■653 ▼aSingular vector decomposition
■653 ▼aSocial media
■653 ▼aNatural language
■690 ▼a0573
■690 ▼a0984
■690 ▼a0723
■71020▼aIndiana University▼bInformatics.
■7730 ▼tDissertations Abstracts International▼g85-02B.
■773 ▼tDissertation Abstract International
■790 ▼a0093
■791 ▼aPh.D.
■792 ▼a2023
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T16934270▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
■980 ▼a202402▼f2024
