The matching engine of exorbyte GmbH is considered one of the fastest and most precise when it comes to searching and matching huge amounts of data. With the new software release, exorbyte brings the excellent matching performance to the Docker container. Scalability and security dominate the new release. Adaptations for simplified integration and administration equip the powerful tool for plug-and-play use as a third-party service in integrated system solutions for efficient data management in a wide range of business areas.
matchmaker 7 simplifies integration into modern system landscapes and enables new scaling options. Starting with version 7, matchmaker is available as a pre-installed ready-to-use Docker image. Individual matching solutions are configured on top of the base image and also delivered to the customer as a ready-to-go image.
All client-server communication is secured via the recognized standard for transport encryption TLS 1.3. Optionally, this can be additionally authorized by a freely configurable API key. Web services use the proven HTTPS protocol for data exchange. The REST interfaces can be secured with a mouse click in the configuration. This means that queries and sensitive index data are only visible to the authorized user group.
The interfaces for the administration of matchmaker 7 have been unified and simplified. Thus, simultaneous work via graphical user interface as well as via command line and JAVA interface is possible on the same matchmaker instance.
New classes for high-performance asynchronous query processing are available for batch processing in all native client interfaces (Java, C#, C++, Python). Automated matching processes are thus completed even faster.
exorbyte keeps pace with the evolution of the open source application development platform. matchmaker 7 now offers a suitable interface for apps programmed for .NET Core environments.
The new indexing method "Street Method" makes the matching of street addresses even more precise, as country-specific prefixes and suffixes such as "-street" or "-place" can now be easily recognized without complex scripting and weighted separately in relation to the actual street name. Thus matchmaker focuses on the relevant part of the address information and eliminates irrelevant hits.
The method is pre-configured for DACH, US or UK/IE addresses and can easily be adapted to other language and address spaces.
The integrated alias generators for the generation of variants, e.g. on name, street or location fields, are even more performant in matchmaker 7 and even easier to use thanks to graphical configuration elements in exTractor.
FlexForm, the flexible multi-field search that enables matching of unstructured information when field mapping is unclear, has been further improved and is now up to 60% faster than before. Thanks to new recall configuration, linking unstructured search (FlexForm) and search with fixed field mapping (FixedForm) is now even easier.
Transparency about query volume: The new matchmaker Query Count Reporting generates a signed report about the query load per session configuration every hour. This guarantees secure and accurate usage-based billing. No queries or other sensitive data are logged in the process.
Improved administration: By redesigning the project directory tree in matchmaker 7.1, configuration, index and log files are more clearly structured and easier to navigate. In addition, the new path completion in the exMinistrator command line makes navigation easier. Index structures are also easier to manage because temporary data and static configuration data have been separated. Log rotation and archiving are now more systematic and consistent across all matchmaker tools.
Improved integration: In addition to the existing REST interfaces, the new matchmaker Recall Bridge now offers the full functionality of the native matchmaker interfaces also for asynchronous queries as a secured REST API.
The redesigned matchmaker installer allows easier selection of components to install to customize the scope of the installation. A streamlined matchmaker image is now also available for Docker. On Windows, the matchmaker start menu has been structured more clearly. matchmaker is now automatically added to the startup path.
The instant update represents a significant improvement of the incremental update. It allows to immediately inject the main elements of the changed data into the search and thus make them immediately searchable.
The best of single-field and multi-field search combined: The user can enter queries in multiple fields with the flexible form search (FlexForm), matchmaker performs the correct mapping to the data dynamically and depending on the best "interpretation" of the query.
Multi-Level-Search is a powerful search tool that makes search queries much more flexible. It allows to search within defined hierarchy levels, e.g. for cities within a people database. Even queries like "find all streets where both a Mr. Schmitz and Mr. Maier live" are now possible.
matchmaker now supports data extraction from XML files. The configuration process is graphically displayed and therefore easy to use. This feature is especially helpful when used with multi-level search.
The Configuration Wizard statistically analyzes the data to be integrated with regard to its structure. Based on this analysis, the wizard makes suggestions for a reasonable matchmaker configuration.
This feature is particularly advantageous for very large data volumes and short update cycles. In addition to the normal incremental update, the "Distributed Build" offers the possibility to distribute the indexing to several processors and/or machines and additionally to partially parallelize it. This accelerates the indexing process.
The new feature automatically extracts color values from images and allows searching for them using color names (,,blue"), color ranges (,,reddish") and RGB or HSV values. Thus, it is now possible to search for images with similar colors, such as all "red" or "reddish" products.
With this extension exLog, the matchmaker log server, can now be connected directly to external monitoring systems like Nagios. Thus log files can be monitored centrally in a monitoring system.
The new tool exPlain analyzes the functioning of algorithms that are difficult to follow, such as Levenshtein, and presents the relationships in graphical form. Designed as a test environment and learning tool, exPlain helps to better understand matchmaker structures and the effects of search configurations.