Context recognition using internet as a knowledge base

• Published in: Journal of intelligent information systems Vol. 29; no. 3; pp. 305 - 327
• Main Authors: Segev, Aviv, Leshno, Moshe, Zviran, Moshe
• Format: Journal Article
• Language: English
• Published: New York Springer Nature B.V 01.12.2007
• Subjects: Algorithms; Artificial intelligence; Automation; Classification; Comparative studies; Data analysis; Human subjects; Information retrieval; Internet; Knowledge; Metadata; Ontology; Recognition
• ISSN: 0925-9902; 1573-7675
• DOI: 10.1007/s10844-006-0015-y

Context recognition is an important component of the common sense knowledge problem, which is one of the key research areas in the field of Artificial Intelligence. The paper develops a model of context recognition using the Internet as a knowledge base. The use of the Internet as a database for context recognition gives a context recognition model immediate access to a nearly infinite amount of data in a multiplicity of fields. Context is represented here as any textual description that is most commonly selected by a set of subjects to describe a given situation. The model input is based on any aspect of the situation that can be translated into text (such as: voice recognition, image recognition, facial expression interpretation, and smell identification). The research model is based on the streaming in text format of information that represents situations--Internet chats, e-mails, Shakespeare plays, or article abstracts. The comparison of the results of the algorithm with the results of human subjects yielded a very high agreement and correlation. The results showed there was no significant difference in the determination of context between the algorithm and the human subjects. [PUBLICATION ABSTRACT]