| 刘昱昕,张延华.基于ANFIS-MBTI的人格类型指标自动检测方法[J].高技术通讯(中文),2025,35(7):734~745 |
| 基于ANFIS-MBTI的人格类型指标自动检测方法 |
| Automatic detection method for personality type indicators based on ANFIS-MBTI |
| |
| DOI：10. 3772 / j. issn. 1002-0470. 2025. 07. 006 |
| 中文关键词: 迈尔斯-布里格斯人格类型指标分类;机器学习;自适应神经模糊推理系统;模糊逻辑 |
| 英文关键词: Myers-Briggs type indicator, machine learning, adaptive-network-based fuzzy inference system, fuzzy logic |
| 基金项目: |
| 作者 | 单位 | | 刘昱昕 | (北京工业大学信息学部 北京 100124) | | 张延华 | |
|
| 摘要点击次数: 341 |
| 全文下载次数: 657 |
| 中文摘要: |
| 迈尔斯-布里格斯人格类型指标分类(Myers-Briggs type indicator,MBTI)测验被认为是预测人格类型最热门和最可靠的方法之一,但传统的问卷调查或专业人士咨询的检测方式在实施过程中面临着高昂的人力和时间成本以及潜在的隐私泄露风险。 针对这类问题,本文提出一种基于自适应神经模糊推理系统( adaptive-network-based fuzzy inference system,ANFIS)的 MBTI 模型(ANFIS-MBTI)。 该模型将深度神经网络与模糊逻辑推理有机融合,使其能够通过自学习和参数优化策略,灵活适应并精准捕捉社交文本数据中隐含的非线性、模糊和不确定性特征,自动识别出分析社交媒体数据集中的用户行为模式,从而揭示其在信息获取、决策制定及行为方式等方面的心理特质和性格特点。实验结果表明,本文构建的 ANFIS-MBTI 模型能够高效而准确地从社交文本中挖掘出 16种不同的 MBTI 人格类型,其多层级特征融合机制使人格分类任务的自动化程度显著提升;同时通过模糊规则约束有效控制人工干预需求与数据隐私风险,为大规模在线人格分析提供了具有可扩展性的创新技术路径。 |
| 英文摘要: |
| Myers-Briggs type indicator (MBTI) is regarded as one of the most popular and reliable methods for predicting personality types. However,traditional detection approaches such as questionnaire surveys or professional consultations face high human and time costs as well as potential privacy leakage risks during implementation. To address these issues,this paper proposes an MBTI classification model (ANFIS-MBTI) based on the adaptive-network-based fuzzy inference system (ANFIS). By organically integrating deep neural networks with fuzzy logic reasoning,the model can flexibly adapt to and accurately capture nonlinear,ambiguous,and uncertain features hidden in social text data through self-learning and parameter optimization strategies. This enables automatic identification of user behavior patterns in social media datasets,thereby revealing psychological traits and personality characteristics in information acquisition,decision-making,and behavioral patterns. Experimental results demonstrate that the proposed ANFIS-MBTI model efficiently and accurately identifies 16 distinct MBTI personality types from social texts. Its multilevel feature fusion mechanism significantly enhances the automation level of personality classification tasks,while fuzzy rule constraints effectively control manual intervention requirements and data privacy risks,providing a scalable innovative technical pathway for large-scale online personality analysis. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
