対応プラットフォーム
Windows, Linux, Solaris, AIX, HPUX, MacOS 等
Document Exploitation (DOCEX) Triage and Analysis
Software to find meaningful data from real-world text data
Before analysts can “connect the dots,” those key “dots” must be recognized and rendered understandable. They may be obscured as Arabic chat alphabet, written in a foreign language, locked up in a cellphone, or simply lost among the mass of captured information.
Governments Rely on Basis Technology
Governments around the world and agencies of the U.S. intelligence community, the Department of Defense, the FBI and others use our products for intelligence triage in critical languages—Arabic, Chinese, Dari, Farsi, Pashto, and Urdu. Our text analytics and name technologies serve vital functions in monitoring and sifting open source intelligence; correlating names against watch lists and between disparate reports; and boosting the productivity of scarce analysts and linguists.
Unlock DOCEX in All Critical Languages
Basis Technology’s Rosette text analytics platform can process volumes of data automatically, identifying 55 languages, extracting names of people and places, and annotating names in foreign documents with English. The output of Rosette feeds into visualization and link analysis applications or alerting systems.
Software Components to Build a DOCEX Screening Platform
Lightweight, fast, and scalable, our DOCEX components can be used as the basis for a new DOCEX application to quickly screen for high priority documents to be exploited immediately. Alternatively, our DOCEX components are easily integrated into pre-existing systems to provide the exact combination of functionality and language support required, without slowing overall performance.
Entity Extraction and Name Matching to Power Link Analysis and Alerting Systems
Our DOCEX components provide automatic entity extraction (people, places, organizations, etc.) from data and find matching names. These data are the raw materials for link analysis and alerting systems, triggered to key names and concepts. Our entity extractor operates in 18 languages with options for users to add entities. The name matching component finds all references to a single person despite variant name spellings, typos, or names written in other languages, and operates in critical languages such as English, Arabic, Chinese, Dari, and Pashto.
All the Components for a DOCEX Pipeline
Rosette Entity Extractor automatically flags the names of people and places in documents, among other entities. Feeding foreign script names through Rosette Name Translator turns them into readable English for quick gisting when a linguist expert is not available. Similarly, when plugged into a search application, Rosette Name Indexer takes names searched in English, and finds all documents with that name even if it is spelled differently or written in Arabic, Korean, Persian, Pashto, and other languages.
A combination of Rosette Entity Extractor and Rosette Name Translator
outputs name translations superior to commercial machine translation systems,
first extracting names from the text and then translating them into
English.
For More Information
Learn More
For more information about how Rosette can be integrated into an existing system or built into a custom solution, please explore: