Content and Sentiment Analysis on Online Social Networks (OSNs)

Fatta, Davide Di; Musotto, Roberto

doi:10.1007/978-3-319-54499-1_5

Cited by 4 publications

(2 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the last years, text classification for analyzing blog posts, Twitter or Facebook data became increasingly popular. For social media, predefined categories are often related to some task of sentiment analysis, which aims to measure a positive or negative affect of authors toward certain issues, events, or persons (e.g., Di Fatta & Musotto, 2017; Jungherr, Schoen, Posegga, & Jürgens, 2016; Lewis, Zamith, & Hermida, 2013; Su et al, 2016). Despite its innovative potential, so far there is only little use of supervised learning for CA beyond sentiment classification in social media (Denny & Spirling, 2017).…”

mentioning

confidence: 99%

Proportional Classification Revisited: Automatic Content Analysis of Political Manifestos Using Active Learning

Wiedemann

2018

Social Science Computer Review

View full text Add to dashboard Cite

Supervised machine learning is a promising methodological innovation for content analysis (CA) to approach the challenge of ever-growing amounts of text in the digital era. Social scientists have pointed to accurate measurement of category proportions and trends in large collections as their primary goal. Proportional classification, for example, allows for time-series analysis of diachronic data sets or correlation of categories with text-external covariates. We evaluate the performance of two common approaches for this goal: a method based on regression analysis with feature profiles from entire collections and a method aggregating classifier decisions for individual documents. For both, we observed a significant negative effect on classification performance due to the uneven distribution of characteristic language structures within the text collection. For proportional classification, this poses considerable problems. To fix this problem, we propose a workflow of active learning, which alternates between machine learning and human coding. Results from experiments with empirical data (political manifestos) demonstrate that active learning enables researchers to create training sets for automatic CA efficiently, reliably, and with high accuracy for the desired goal while retaining control over the automatic process.

show abstract

mentioning

confidence: 99%

Proportional Classification Revisited: Automatic Content Analysis of Political Manifestos Using Active Learning

Wiedemann

2018

Social Science Computer Review

View full text Add to dashboard Cite

show abstract

“…This definition has been criticized by Berger and Luckmann (1966) who argued that it is not possible to produce totally objective results, because the analysis will be always influenced by the interpreter of data. Nevertheless, Krippendorff’s conceptualization is the most widely accepted in the literature that recently showed a boom in content analysis usage (Bernard, 2011; Di Fatta et al , 2016; Di Fatta and Musotto, 2017).…”

Section: Methodsmentioning

confidence: 99%

Modelling multiples identity types through agency: Part 3 – mindsets and the Trump election

Fatta

Yolles

2017

Self Cite

View full text Add to dashboard Cite

Purpose Building on theory in Part 2 of this paper, a relationship is developed between the strategic multiple identities considered there. Personality analytic pathologies arise when these identities are not consistent. This theory is then examined using the mindset agency theory (MAT) developed in Part 2 of the paper. Two classes of MAT models exist: a three-trait (MAT3T) and a five-trait (MAT5T). The former centres on personality traits, while the latter includes traits that are external to the personality. These are then applied to a case study of Donald Trump’s US election campaign. Design/methodology/approach By applying MAT3T and MAT5T to the Trump election campaign, personal and public identities are analysed using content analysis of his narratives. Findings Of the strategic identities, data can be accessed for two, and measured qualitatively using mindset theory, these indicating the likelihood of a personality with pathologies. It is found that Trump MAT3T and MAT5T take different values, suggesting that he has an analytical pathology in his political agency. Originality/value There is not currently any coherent dynamic theory of multiple identities able to provide measures indicative of personality pathologies.

show abstract

Modelling identity types through agency: part 2, personal identity and mindsets

Fatta

Yolles

2017

Self Cite

View full text Add to dashboard Cite

Purpose The purpose of this paper is to evaluate the personal identity of Donald Trump in the US presidential election using the mindset agency theory framework and content analysis. Design/methodology/approach The qualitative evaluation of identity type is determined by the personality mindset agency theory (PMAT). This measures qualitatively by assigning a type to the personal identity. The methods being adopted are content analysis, and a coding frame is constructed that arises from the key words defined in PMAT. Findings Using PMAT, the authors determine that Trump’s personal identity is of the type hierarchical popularism (HP), from which behavioural patterns are derived, supposing that this is consistent with his public identity type measured using agency MAT (AMAT), which will be assessed in part 3 of this paper. Originality/value Appropriate image management can be used in an attempt to hide problematic purely self-interest aspects of a personality. This paper shows that it is possible to evaluate personality mindsets using content analysis. In a later paper, exploration of agency mindsets will occur that is indicative of the potential for behaviour.

show abstract

Content and Sentiment Analysis on Online Social Networks (OSNs)

Cited by 4 publications

References 58 publications

Proportional Classification Revisited: Automatic Content Analysis of Political Manifestos Using Active Learning

Proportional Classification Revisited: Automatic Content Analysis of Political Manifestos Using Active Learning

Modelling multiples identity types through agency: Part 3 – mindsets and the Trump election

Modelling identity types through agency: part 2, personal identity and mindsets

Contact Info

Product

Resources

About