Search results for "topic model"
showing 10 items of 23 documents
A two-stage LDA algorithm for ranking induced topic readability
2022
Probabilistic topic models, such as LDA, are standard text analysis algorithms that provide predictive and latent topic representation for a corpus. However, due to the unsupervised training process, it is difficult to verify the assumption that the latent space discovered by these models is generally meaningful and valuable. This paper introduces a two-stage LDA algorithm to estimate latent topics in text documents and use readability scores to link the identified topics to a linguistically motivated latent structure. We define a new interpretative tool called induced topic readability, which is used to rank topics from the one with the most complex linguistic structure to the one with the…
Comparison of MeSH terms and KeyWords Plus terms for more accurate classification in medical research fields. A case study in cannabis research
2021
Abstract KeyWords Plus and Medical Subject Headings (MeSH) are widely used in bibliometric studies for topic mapping. The objective of this study is to compare the two description systems in documents about cannabis research to find the concordance between systems and establish whether there is neutrality in topic mapping. A total of 25,593 articles from 1970 to 2019 were drawn from Web of Science's Core Collection and Medline and analyzed. The tidytext library, Zipf's law, topic modeling tools, the contingency coefficient, Cramer's V, and Cohen's kappa were used. The results included 10,107 MeSH terms and 28,870 KeyWords Plus terms. The Zipf distribution of the terms was different for each…
Analysing Tourist Destination Image through Topic Modeling
2019
Topic modeling has become one of the most used methods to analyse textual data, proving able to “discover” hidden dimensions (topics) which characterise a corpus. This methodology can be used fruitfully to analyse complex phenomena like tourist destination image. With this aim in mind, this paper discusses the use of topic modeling over TV commercials which have been broadcast by four of the major cruise lines operating in Italy in recent years.
Automated Content Analysis of Destination Image: a Case Study
2020
Automated content analysis has become one of the most used approaches to extract “hidden” dimensions from text corpora over the last years. One of the data analysis techniques belonging to this approach is topic modeling, which can be fruitfully used to analyse complex phenomena like tourist destination image. With this aim in mind, this paper discusses the use of topic modeling to identify the main components of the image of cruise holidays spread through a specific type of visual text, i.e. the Television commercial. In order to achieve this goal, the paper presents the methodology and main results of a study carried out over a sample of TV commercials, which have recently been broadcast …
Staying at the front line of literature: How can topic modelling help researchers follow recent studies?
2021
Staying at the front line in learning research is challenging because many fields are rapidly developing. One such field is research on the temporal aspects of computer-supported collaborative learning (CSCL). To obtain an overview of these fields, systematic literature reviews can capture patterns of existing research. However, conducting systematic literature reviews is time-consuming and do not reveal future developments in the field. This study proposes a machine learning method based on topic modelling that takes articles from a systematic literature review on the temporal aspects of CSCL (49 original articles published before 2019) as a starting point to describe the most recent devel…
Exploring the challenges of remote work on Twitter users’ sentiments: From digital technology development to a post-pandemic era
2022
The boost in the use and development of technology, spurred by COVID-19 pandemic and its consequences, has sped up the adoption of new technologies and digital platforms in companies. Specifically, companies have been forced to change their organizational and work structures. In this context, the present study aims to identify the main opportunities and challenges for remote work through the use of digital technologies and platforms based on the analysis of user-generated content (UGC) in Twitter. Using computer-aided text analysis (CATA) and natural language processing (NLP), in this study, we conduct a sentiment analysis developed with Textblob, which works with machine learning. We then …
Isotopic Anthropology of Rural German Medieval Diet: Intra- and Inter-population Variability
2016
This study investigates the diet of an eleventh century CE parish community located in northwestern Germany. We assessed the isotopic compositions of human (n = 24) and faunal (n = 17) bone collagen (δ 13Ccol, δ 15Ncol) and human structural carbonate (δ 13Csc) using skeletal material recovered from the Dalheim cemetery. Traditional interpretation of the isotopic data indicates that Dalheim residents likely relied on a C3 plant-based diet and consumed some terrestrial animal products without evidence of marine resource input in the diet. Bivariate and multivariate models used as an additional means to assess diet indicate minor consumption of C4 plant foods in this community. The multivariat…
Commenting on poverty online : A corpus-assisted discourse study of the Suomi24 forum
2020
This paper brings new insight to poverty and social exclusion through an analysis of how poverty-related issues are commented on in the largest online discussion forum in Finland: Suomi24 (‘Finland24’). For data, we use 32,407 posts published in the forum in 2014 that contain the word köyhä (‘poor’) or a predefined semantically similar word. We apply the Corpus-Assisted Discourse Studies (CADS) method, which combines quantitative methods and qualitative discourse analysis. This methodological solution allows us to analyse both large-scale tendencies and detailed expressions and nuances on how poverty is discussed. The quantitative analysis is conducted with topic modelling, an unsupervised …
Examining Competing Entrepreneurial Concerns in a Social Question and Answer (SQA) Platform
2021
This study aims to determine the competing concerns of people interested in startup development and entrepreneurship by using topic modeling and sentiment analysis on a social question-and-answer (SQA) website. Understanding the underlying concerns of startup entrepreneurs is critical to society and economic growth. Therefore, greater scientific support for entrepreneurship remains necessary, including data mining from virtual social communities. In this study, an SQA platform was used to identify the sentiment of thirty concerns of people interested in startup entrepreneurship. Based on topic modeling and sentiment analysis of 18819 inquiries in various forums on an SQA, we identified addi…
H∞ fuzzy control of DC-DC converters with input constraint
2012
Published version of an article in the journal: Mathematical Problems in Engineering. Also available from the publisher at: http://dx.doi.org/10.1155/2012/973082 Open access This paper proposes a method for designing H∞ fuzzy control of DC-DC converters under actuator saturation. Because linear control design methods do not take into account the nonlinearity of the system, a T-S fuzzy model and a controller design approach is used. The designed control not only handles the external disturbance but also the saturation of duty cycle. The input constraint is first transformed into a symmetric saturation which is represented by a polytopic model. Stabilization conditions for the H∞ state feedba…