Project Description:
Semantic recognition distinguishes speech recognition, which is closer to artificial intelligence. The semantic recognition at this stage is basically realized by the traditional search technology based on mass database, and its essence is: edit of artificial knowledge points + word segmentation + keyword matching. What the engine does is memorize the whole memory of the book, but in reality users need to answer predictable or unpredictable problems that need to be flexibly applied on the basis of automatic extraction, splitting and recording of knowledge points, The method has a lot of limitations, requiring a new generation of semantic recognition technology that uses 'knowledge-formatting'.
DOONE uses proprietary semantics elements of the SGK (Sematic Guide Keywords), which are proprietary intellectual property rights, to guide the key elemental algorithms, formatting knowledge flexibly and in a regular way, combining 'key elements', 'semantic elements' Neural network ',' transmission path 'and other neural network characteristics match naturally, and through the algorithm of clustering such as degenerate algorithm, greedy algorithm, BP algorithm and so on, so that the semantics can be realized even if the customer database is not perfect and regular Elements lead the software engineering of key elements DOONE's software products have been applied to the business of e-government, Audi automobile marketing and human resources marketing of Shanghai Wujiaochang Street, Bridge Street, Shanghai Wanda Information and China-Chile Care Communication.
Purpose of financing: One is to improve the SGK semantic elements guide the key elements of the algorithm code optimization; the second is to choose a giants to develop an industry-leading customer service interaction system;
Project volume: This round of financing 10 million yuan
Contact E-mail: lionblaze@yeah.net; www.doonesoft.com