Recently, TRS Hybase Search (Vector) Database and TRS HaiWen Natural Language Processing Engine independently developed by TRS have passed the evaluation requirements of "software independent control" by an authoritative organization. The evaluation results show that the overall source code autonomy rate and core module source code autonomy rate of the two products exceed 80%, reaching the independent controllable B+ level capability.

TRS Hybasel Search (Vector) Database is one of the few home-made self-developed software in the domestic market that meets both search database standards and vector database standards, which can meet the domestic replacement needs of related products. On the basis of inverted index, the system introduces high-performance HNSW vector index, which not only supports common text retrieval, pinyin retrieval, synonym retrieval, etc., but also can be used with large models (deep learning engines) to realize semantic search, image search, cross-modal search and fusion search, etc. It can be widely used in enterprise search, retrieval enhancement generation (RAG) and intelligent question answering.

TRS HaiWen Natural Language Processing Engine (Deep text mining software TRS DL-CKM) is a new generation of NLP engine based on deep learning. It absorbs TRS's technical accumulation in the fields of natural language processing and information retrieval for many years. It focuses on core technologies such as deep learning and knowledge map, relies on TRS intelligent semantic intelligent technology platform and other technical achievements, and is oriented to application scenarios such as intelligent patents, intelligent public security, intelligent government affairs, intelligent finance, open source intelligence analysis, etc. Intelligent empowerment of users 'business applications.

In the future, TRS will continue to adhere to independent research and development of core technologies to build a solid defense line for national security.