FAVORITE SITEMAP

home ＞ Development Information ＞ Semantic Recognition

Voice Technology Semantic Recognition Morphological Analysis Product Development Release Information

Many talking tools including “Artificial Intelligence (AI)” and “Artificial Unintelligence” have been proposed. However, an exclusive Knowledge Base (KB) or Database (DB) is provided for these tools and users are unable to create or edit the knowledge base anew. In contrast, it is the ‘Semantic Recognition Feature’ introduced into CAIWA that has enabled realization of Artificial Intelligence that enables general users to create and edit the knowledge base. Before discussing Semantic Recognition, let us explain “Conversational Contents” handled in CAIWA here. The contents of conversations are classified into; ①　Q&A or FAQ where responses are given to individual questions by users respectively　②　A scenario type similar to an instructor, being questioned, talking one-sidedly ③　A chat type where the first two types are combined. In regard to ①　that deals with Q&A or FAQ, it is necessary to expand expression expansion installed with semantic correlations between words so that contents of questions are related to appropriate response sentences. In the scenario-type of ②, “Semantic Analysis” is necessitated in order to properly induce answers to the questions from users with unity between CAIWA and response sentences. These expression features such as ‘expansion search’ and ‘semantic analyses are the fundamental features in ‘semantic recognition’ of CAIWA. In order to replenish semantic recognition features substantial semantic recognition dictionaries are mounted in CAIWA, such as thesauruses for semantic analyses and ontological environments for expression expansion searches. CAIWA is not different from other Artificial Intelligence in that natural language processing technology based on morphophonemic analyses is applied against question sentences of users and response sentences.　In CAIWA, however, introduction of such substantial semantic recognition dictionaries has resulted in enabling ordinary users to create and edit the database with ease. It can be expected that creating and editing artificial intelligence can be transferred from the hands of expertise to those of ordinary users with an easily created and edited knowledge base. It can be said that the fact that replenishing semantic recognition features has made it possible for anyone to create and edit artificial intelligence may be the greatest feature of CAIWA provided by PtoPA.