An Improved Study of Multilevel Semantic Network Visualization for Analyzing Sentiment Word of Movie Review Data

Ha, Hyoji; Han, Hyunwoo; Mun, Seongmin; Bae, Sungyun; Lee, Jihye; Lee, Kyungwon

doi:10.3390/app9122419

Cited by 21 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The most frequently used POS-Tag is Penn-POS-Tag, and research [4] proposes to select SWN-POS-Tag for the SWN dictionary with the following comparison [10]. Conversion POS between SWN and Penn is shown in table (7).…”

Section: Feature Extraction Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Sentiment Classification for Film Reviews by Reducing Additional Introduced Sentiment Bias

Effendi

Sibaroni

2021

RESTI

View full text Add to dashboard Cite

Film business and its individual reviews cannot be separated and film review sites such as IMDb is a credible source of reviews posted in public forums. With IMDb site reviews being unstructured and bias-heavy, classification methods by reducing additional sentiment bias is needed to create a balanced classification with lower polarity bias. Elimination of additional sentiment bias will improve the model as polarity is defined by non-bias method, resulting in models correctly defined which sequences of words is either positive or negative. This research limits the dataset by 50.000 rows of randomly extracted reviews from the IMDb website using dataset preparation methods such as Preprocessing, POS-Tagging, and Word Embeddings. Then preprocessed data is used in classification methods such as ANN, SWN, and SO-Cal. This paper also used bias processing methods such as Hyperparameter Tuning and BPM, with outputs evaluated using Accuracy and PBR metrics. This research yields 77.39 % for ANN, 66.32% for BPM, 75.6% for SO-Cal, and 76.26% for Hybrid classification. Best PBR resulted in two lexicon-based methods on 0.0009 for BPM, and 0.00006 for SO-Cal. More advanced model configuration in ANN can improve the model, and much complex lexicon models will be a future in the research topic.

show abstract

Section: Feature Extraction Methodsmentioning

confidence: 99%

“…In the dataset In study [6] with different datasets, sentiment analysis done using the Lexicon-based method has a pretty good performance with the SVM Classifier with an accuracy of ¬83.27%. Multilevel Semantic Network was suggested by Research [7] for sentiment analysis method with similar data with a maximum accuracy of 74.2%.…”

Section: Introductionmentioning

confidence: 99%

Sentiment Classification for Film Reviews by Reducing Additional Introduced Sentiment Bias

Effendi

Sibaroni

2021

RESTI

View full text Add to dashboard Cite

show abstract

“…The present study was exploratory rather than hypothesis testing [3,37]. Therefore, our main aim was not to explore the impact of certain variables, but to discover them and illustrate how they can be analyzed in future research using the proposed methods [37,38]. Therefore, in the present study, we used a qualitative and exploratory research methodology [13,16].…”

Section: Methodsmentioning

confidence: 99%

Comparing Data-Driven Methods for Extracting Knowledge from User Generated Content

Saura

Reyes-Menéndez

Filipe

2019

Journal of Open Innovation: Technology, Market, and Complexity

View full text Add to dashboard Cite

This study aimed to compare two techniques of business knowledge extraction for the identification of insights related to the improvement of digital marketing strategies on a sample of 15,731 tweets. The sample was extracted from user generated content (UGC) from Twitter using two methods based on knowledge extraction techniques for business. In Method 1, an algorithm to detect communities in complex networks was applied; this algorithm, in which we applied data visualization techniques for complex networks analysis, used the modularity of nodes to discover topics. In Method 2, a three-phase process was developed for knowledge extraction that included the application of a latent Dirichlet allocation (LDA) model, a sentiment analysis (SA) that works with machine learning, and a data text mining (DTM) analysis technique. Finally, we compared the results of each of the two techniques to see whether or not the results yielded by these two methods regarding the analysis of companies’ digital marketing strategies were mutually complementary.

show abstract

“…In order to study the correlation between network big data and film time-series data, it is necessary to carry out a series of data collection aiming at the relevant characteristics of films [17]. In order to more accurately study film time-series data, this paper adopts relevant data from Facebook social platform for analysis [18,19]. e data of this platform can further ensure the accuracy of experimental data [20].…”

Section: Test Data Acquisitionmentioning

confidence: 99%

Correlation Analysis of Network Big Data and Film Time-Series Data Based on Machine Learning Algorithm

Li¹,

Xia

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

To expand the application of machine learning in movie data, in order to explore the correlation between network big data and film time-series data, based on the machine learning algorithm, the correlation and multifractal characteristics of happiness index (HI) and film box office (BO) were studied and described by introducing multifractal crossover method. On this basis, some indicators are introduced to optimize the neural network model so that the optimization model can describe and predict the box office and other related information well. The results show that the critical values of the happiness index and box office show a linear change trend with the increase of freedom, and the corresponding change curves of the happiness index and box office show obvious nonlinear characteristics, which can be divided into slow increase stage, steady increase stage, and approximately gentle stage. With the increase of iteration parameter q value, the change trend of the long-term and short-term curves of the generalized Hurst function is basically the same, and the difference between the two is getting smaller and smaller, while the difference between the two curves is getting bigger and bigger with the increase of q value of Renyi function. The changing trend of the dynamic Hurst index in the sliding window period all shows that it first rises rapidly to a certain value, then fluctuates rapidly with the increase of time, then drops rapidly to a constant value, and finally continues to show repeated small range fluctuation. Under the influence of time-series parameter α, the original sequence changes the most, the replacement sequence changes the medium, and the corresponding rearrangement sequence changes the least. The overall distribution of box office prediction data conforms to the characteristics of linear variation. The prediction index of the optimized HI-LSTM (Happiness Index-Long term short term memory neural network) model is higher in the box office, indicating that the model has better performance in describing and predicting the box office. This study can provide a theoretical basis for the correlation study of network big data and film data.

show abstract

An Improved Study of Multilevel Semantic Network Visualization for Analyzing Sentiment Word of Movie Review Data

Cited by 21 publications

References 19 publications

Sentiment Classification for Film Reviews by Reducing Additional Introduced Sentiment Bias

Sentiment Classification for Film Reviews by Reducing Additional Introduced Sentiment Bias

Comparing Data-Driven Methods for Extracting Knowledge from User Generated Content

Correlation Analysis of Network Big Data and Film Time-Series Data Based on Machine Learning Algorithm

Contact Info

Product

Resources

About