KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/kmitl/1

Browse

Settings

Subject: search.filters.subject.Natural language processing

Search Results

Now showing 1 - 4 of 4
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Fast Hybrid Approach for Thai News Summarization
    (2024-07-01)
    Jearanaitanakij, Kietikul
    ;
    Boonpong, Suratan
    ;
    Teainnagrm, Kirttiphoom
    ;
    Thonglor, Thanakrit
    ;
    Kullawan, Tiwat
    News summarization presents a significant challenge in Natural Language Processing (NLP). Lengthy news articles not only consume valuable time but also lead to confusion regarding key points. The ideal news summarization should swiftly produce a succinct summary while retaining the essence of the information conveyed by the news writer. While intelligent chatbots like ChatGPT and Gemini offer user-friendly text summarization, their embedded Large Language Model (LLM) cannot be downloaded for private use. Moreover, implementing them in a business process can be expensive, both in terms of pay-per-use costs and response time. The objective of this research is to develop a private Thai news summarization that effectively extracts sentences encapsulating the main idea and abstractly summarizes them. The proposed model consists of two components. The first extracts a contiguous region containing important sentences using the TextRank algorithm, while the second employs the finetuned mBART as an LLM to generate the abstractive summary from the previously extracted sentences. In other words, the proposed model extracts an important news region before passing it to mBART. This approach produces a news summary with key information and a syntactic style akin to the natural Thai language. We evaluate the summarization quality by ROUGE scores and BERTScore (precision, recall, and F1-score). The evaluation metrics Experimental results on the ThaiSum dataset show relatively high ROUGE scores and BERTScore for the proposed model compared to most of the other approaches. Furthermore, it significantly reduces the runtime, keeping it within a reasonable limit.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Comparison of Evaluation Metrics for Short Story Generation
    (2023-01-01)
    Netisopakul, Ponrudee
    ;
    Taoto, Usanisa
    This study aimed to analyze the correlation among different automatic evaluation metrics for short story generation. In the study, texts were generated from short stories using different language models: the N-gram model, the Continuous Bag-of-Word (CBOW) model, the Gated recurrent unit (GRU) model and the Generative Pre-trained Transformer 2 (GPT-2) model. All models were trained on short Aesop's fables. The quality of the generated text was measured with various metrics: Perplexity, BLEU score, the number of grammatical errors, Self-BLEU score, ROUGE score, BERTScore, and Word Mover's Distance (WMD). The resulting correlation analysis of the evaluation metrics revealed four groups of correlated metrics. Firstly, perplexity and grammatical errors were moderately correlated. Secondly, BLEU, ROUGE and BERTScore were highly correlated. Next, WMD was negatively correlated with BLEU, ROUGE and BERTScore. On the other hand, Self-BLEU, which measures text diversity within the model, did not correlate with the other metrics. In conclusion, to evaluate text generation, a combination of various metrics should be used to measure different aspects of the generated text.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Thai stock news classification based on price changes and sentiments
    (2022-01-01)
    Netisopakul, Ponrudee
    ;
    Saewong, Woranun
    This research investigates the daily stock news influences toward a company's stock price direction in the Stock Exchange of Thailand. First, machine learning's text classification methods, namely, naïve Bayes, decision tree, random forest, support vector machine, and the three-layer and the five-layer backpropagation neural networks, are applied to predict the stock price directions using stock news collected during the year 2018. Then, the stock news sentiment is incorporated to help improve the prediction accuracy. Last, a meaningful grouping of stock news is carried out to further improve the direction prediction. The testing dataset collected from January to March 2019 stock news are used for model evaluations. The best accuracy obtained from the baseline dataset using stock news only is 78.6%. When dataset is augmented with sentiments and grouped, the best accuracy increases to 90.6%.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Cross domain sentiment classification of Thai reviews using co-train model
    (2019-01-01)
    Boonpetch, Warakorn
    ;
    Chitsobhuk, Orachat
    Online reviews are significant sources of information, which is useful for supporting customer and entrepreneur decision in terms of product and service satisfaction analysis. Online reviews containing feedback from various domains makes it difficult to analyze and classify all comments at once. The proposed technique analyses the cross-domain Thai review data using a co-train machine learning model. The co-train model consists of multiple single domain specific models followed by refinement analysis for the final sentiment classification. This allows for full flexibility in training of each individual domain, which can lessen the limitation on training complexity due to simple training on single domain. The experiments have been conducted on Wongnai restaurant domain and IMDB movie domain data. Our co-train model can achieve the highest average accuracy of 86.10 percent for cross-domain sentiment classification with approximately 38 seconds processing time.