From mining to meaning: How B2B marketers can leverage text to inform strategy

Tsao, Hsiu‐Yuan; Campbell, Colin; Sands, Sean; Mavrommatis, Alexis

doi:10.1016/j.indmarman.2022.08.007

Cited by 4 publications

(5 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We adopt two topics, From Text to Action Grid and Scale-Directed Text Analysis, extract the content from the research paper "From Mining to Meaning" [11], and for those questions, ground answers, and generated content from LLM ChatGpt4 and LLM + DSK, please refer to Appendix A. In the following, we attempt to compare the BERTScore calculated from the retrieved chunks from LLM ChatGpt4 and our domain knowledge generated from the domain-specific LM with ground answers.…”

Section: Cases 3 Andmentioning

confidence: 99%

Empowering Large Language Models to Leverage Domain-Specific Knowledge in E-Learning

Lu,

Lin,

Tsao

2024

Applied Sciences

Self Cite

View full text Add to dashboard Cite

Large language models (LLMs) have demonstrated remarkable capabilities in various natural language processing tasks. However, their performance in domain-specific contexts, such as E-learning, is hindered by the lack of specific domain knowledge. This paper adopts a novel approach of retrieval augment generation to empower LLMs with domain-specific knowledge in the field of E-learning. The approach leverages external knowledge sources, such as E-learning lectures or research papers, to enhance the LLM’s understanding and generation capabilities. Experimental evaluations demonstrate the effectiveness and superiority of our approach compared to existing methods in capturing and generating E-learning-specific information.

show abstract

Section: Cases 3 Andmentioning

confidence: 99%

Empowering Large Language Models to Leverage Domain-Specific Knowledge in E-Learning

Lu,

Lin,

Tsao

2024

Applied Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, the frequency indicates the extent of attention instead of the extent of affection (preference or valence) [7]. Hence, this study applied the sentiment analysis of those sentences to obtain the affection of the aspect instead of adopting the TF-IDF score to predict or compare the questionnaire ratings [10]. Past research adopted the TF-IDF score or manual labeling to compare or predict the questionnaire ratings, which is different.…”

Section: Categorization and Sentiment Analysis Of Textual Datamentioning

confidence: 99%

“…The methods of personality measurement regarding sentiment are summarized as follows. For detailed theories and verification methods, please refer to the scale-directed text analysis (SDTA) developed by scholars [9,10]. R and PHP languages are used to develop programs to convert qualitative text content analysis into quantitative marketing scale scores based on the existing marketing scale (This study uses two word datasets, respectively, the AFINN sentiment lexicon (http://www2.imm.dtu.dk/pubdb/pubs/6010-full.html, accessed on 10 November 2022) and MBTI personality as developed by filtering the higher score of TF-IDF as described in a previous paragraph.…”

Section: Categorization and Sentiment Analysis Of Textual Datamentioning

confidence: 99%

“…First, this study will attempt to extract words based on the TF-IDF scores to generate a dictionary of customer personality traits, as most past research has created. Second, we argue that the core technique of TF-IDF is to count the frequency of words, which is based on the extent of attention rather than the extent of affection (preference or valence) [10]. Hence, this study will consider words associated with a specific aspect and apply sentiment analysis to examine sentences that include words necessary to obtain the affection of an aspect rather than adopting TF-IDF scores to predict or compare the questionnaire ratings [8,11,12].…”

Section: Introductionmentioning

confidence: 99%

“…Finally, we adopt the strategic analysis grid of FTTA (From Text to Action), which is an analysis framework based on an aspect to discover four interactions of attention (frequency) and affection (sentiment) [10] to further explore how consumers with specific personality traits use those featured words in term of the interaction of attention (TF-IDF score) and affection (sentiment score). The final research question in this study is whether people with specific personality traits intensively use specific feature words positively or negatively.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predicting Consumer Personalities from What They Say

Tsao

Lin

et al. 2023

Applied Sciences

Self Cite

View full text Add to dashboard Cite

This study mapped personality based on the newly proposed extraction method from consumers’ textual data and revealed the relevance (attention) and polarity (affection) of words associated with a specific personality trait. Furthermore, we illustrate how unique words are used to predict a consumer’s behavior associated with certain personality traits. In this study, we employed the scales of the Kaggle MBTI Personality dataset to examine the methodology’s effectiveness, extract the personality traits from the textual data into features, and map them into the traits/dimensions of the existing scale. Based on the results obtained in this study, we assert that using the TF-IDF algorithm is a good way to generate a custom dictionary. Furthermore, sentiment scoring with an AI-empowered machine learning algorithm provides useful data to filter and validate more coherent words to understand and, thus, communicate a particular aspect of personality. Finally, we proposed that four situations involving the interaction between attention (frequency) and affection (sentiment) allow us to better understand the consumer and how to use the feature words in terms of the interaction between attention (TF-IDF score) and affection (sentiment score).

show abstract