GRAB Thought Dataset for Consciousness Models

Study of mind and nature of intelligence is widely studied in cognitive science. Also, Artificial Wisdom which redefines the Artificial Wisdom is emerging research area where machine intelligence must collaborates with the constructive behavior and values of humanity. Thinking ability of human beings is recognized as the consciousness. Researchers from different domains like Cognitive Science, Artificial Intelligence, Psychology, Computer Engineering etc. are used to perform experimentations on consciousness or arousal of thoughts. As like as every research, research on consciousness or thought generation requires benchmark dataset using which well-build models can be realized. Since research on consciousness is emerging area for the researchers, benchmark dataset is not available or they are not getting benchmark dataset for their work. Unfortunately, research fraternity is striving hard in this regard. Some researchers are collecting data personally based on the application and performed their research work. But they can’t compare their work due to lack of using standard datasets. By considering this thrust in experimenting and developing computational models for consciousness or thoughts, GRAB Thought Dataset is prepared. The word GRAB is named to the dataset since initially it was created for and applied on Gharde-Ramteke Abhidhamma Based (GRAB) Thought computational model of consciousness. GRAB Thought Dataset consists of 445 samples collected from various respondents by asking three questions by showing the images of different feelings, the images of basic shapes like mango, bitter guard, car, etc., the images of few popular personalities, and popular places. It is prepared in CSV format which can be accessible by any supporting applications like Microsoft Excel, etc. GRAB Thought dataset is suitable for the supervised and unsupervised machine learning techniques like Convolutional Neural Network, K-Nearest Neighbour technique, Support Vector Machines, etc. Primary objective of this database is to apply it for generating consciousness and identifying the different types of thoughts and classifying the mental states. Thoughts are classified into four major classes Wholesome (Good), Unwholesome (Bad), Resultant and Functional and mental states are classified into three classes Ethically Variable Factor, Unwholesome Factor and Beautiful Factor. Uses of these classes are completely optional for the user they can omit fields which are specifying these classes. Principle concern of this dataset is to capture the behavior and interest of the respondents and identify their personality traits or behavioral properties or thoughts or consciousness. This database is standardized by applying data cleaning and maintaining proper variance. Various operations of Natural Language Processing can be performed on this dataset.

在认知科学领域，对心智与智能本质的研究已广为人知。与此同时，人工智能智慧（Artificial Wisdom）这一重新定义人工智能的研究领域正崭露头角，其中机器智能必须与人类的建设性行为和价值观念相协作。人类的思维能力被视为意识。来自认知科学、人工智能、心理学、计算机工程等不同领域的学者们，常进行有关意识或思维觉醒的实验研究。如同所有研究一样，对意识或思维生成的研究同样需要基准数据集，以构建健全的模型。鉴于意识研究是研究人员的新兴领域，基准数据集尚未可得或难以获得。遗憾的是，研究界在此方面正面临重重困难。一些研究者基于应用亲自收集数据，并执行其研究工作。但由于缺乏使用标准数据集，他们无法对其工作进行对比。鉴于实验与开发关于意识或思维的计算模型的需求，GRAB思维数据集应运而生。GRAB这一名称源于该数据集最初是为Gharde-Ramteke Abhidhamma Based（GRAB）思维计算模型而创建并应用的。GRAB思维数据集由445个样本组成，这些样本通过展示不同情感的图像、基本形状如芒果、苦楝、汽车等、少数知名人物的图像以及著名场所的图像，从各种受访者中收集而来。该数据集以CSV格式准备，可被任何支持的应用程序如Microsoft Excel等访问。GRAB思维数据集适用于监督学习和无监督学习技术，如卷积神经网络、K-最近邻技术、支持向量机等。此数据库的主要目标是应用于生成意识、识别不同类型的思维以及分类心理状态。思维被分为四大类：善性（良好）、恶性（不良）、结果性和功能性，而心理状态则被分为三类：伦理可变因素、恶性因素和美善因素。用户对这些类别的使用完全可选，他们可以省略指定这些类别的字段。本数据集的核心关注点在于捕捉受访者的行为和兴趣，并识别其人格特质或行为属性或思维或意识。通过对数据进行清洗和维护适当的变异性，该数据库已实现标准化。可以在该数据集上执行各种自然语言处理操作。