A Muse Search Application is an individual instance that runs inside Muse at a customer site or at a group of sites to provide searching or harvesting of resources. It groups a set of modules depending on customer’s desired functionality like:
- Works in scheduled, automated mode e.g. Harvest. Works in on-demand mode e.g. Federated Search;
- Real-time search of disparate multiple Data Sources at once;
- Content returned normalized in format and structure to a single data model;
- Source specific authentication, searching, extraction;
- Results Processing, Deduplication, Cluster, Enrichment, Ranking, etc.;
- On-the-fly Semantic Analysis;
- Output formatting for users, client systems;
- Free access, registered access, subscription access, membership access, employee access;
- Single sign-on;
- Fully customizable simple (Google like) or complex interface.
Multiple Content Sources, Same Search Query, Single Integrated List
It is known that Federated Search appeared as a need to search simultaneously multiple content sources. Federated Search systems search all these sources at a time and present the information to the user in a normalized format. Thus, the user’s effort is much smaller than interpreting these results presented by each source in its own format. The efforts are much lowered – instead of using many browsers accessing individual web interfaces, clients for SOAP and Z39.50 databases or Telnet terminals, the user just accesses a Muse Search Application. Another aspect is that the user doesn’t have to know the syntax for various search queries of each content source – only the search query syntax of the Muse Federated Search.
Query Suggestion and Typeahead, Related Queries
The Query Suggestion features provide suggestions as users type in the search box. As users type within the search box, the suggestion algorithm predicts and displays search queries based on other users' search activities. The Typeahead feature reduces the number of keystrokes needed to type the full term.
Related Queries provides users with query terms that are related to the one used for running the current search. The related queries are displayed in a dedicated interface widget, clicking on any of them will launch a new search.
Query Translation And Query Spell Check
Queries entered in a different language than the current selected language of the interface, are automatically translated into the interface language and the user has the possibility to run a new search with the translated value.
A Spell Check functionality is available to show misspellings in the user query. Each query term is compared with a known list of correctly spelled words (i.e. a dictionary) and lets the user know the closest alternative to his input.
Alternative Text Input Methods
The Muse Search Application features Virtual Keyboard and Speech-to-Text so users can enter more easily texts without the need of the physical keys. These input methods are mostly useful when using public computers, because they are more secure than using the physical keyboard.
Many times users want to have more from a search. Not all content sources present all the information available on a book for example. By our enrichment modules, users may have detailed information retrieved and presented together with the information from the main source, with no supplementary efforts. Many enrichment services can be used simultaneously to have thus more chances to deliver the data that is looked for.
It is known that many search services deliver exactly the same information indexed from various content sources. Why lose time and search many sources for something that you need and then see that a large percent of the information is the same for all? Just use the Muse Search Application which does a smart deduplication for you.
Ranking is a fundamental problem in information retrieval and it attempts to measure how relevant results are to a particular search query. The advanced ranking algorithms we use, together with the multiple taxonomies for various entities, make our ranking one of the best. More, alphabets can be edited via dedicated administrator consoles sections.
Multiple Results Display
No matter what the content sources can display, Muse Search Applications can extend these capabilities when showing the results retrieved. Results can be displayed in full, brief, enriched formats, each being customized to display only the fields wanted. Besides display formatting, users can sort results by any retrieved field, either ascending or descending, or by relevance. There even is the possibility to force the results from some sources to be displayed before others, no matter which came first. There are also other display settings such as the number of results per page, results per source.
Search Related Widgets
All search actions are extended by a set of configurable widgets that present additional related data:
- Refine: relevant terms extracted from the results are grouped by Authors, Dates, Content Types, Languages, Locations, Products, Companies. Users can easily find the associated results by filtering on the terms;
- Topics: groups and visually displays clusters of data to identify emerging trends, topics, and information gaps;
- Extra information for the searched term or phrase using well known data sources such as Wikipedia or Dictionary.
- Sources grouped by their publishers for filtering capabilities;
- All the images the search retrieved are grouped in a carousel layout for an easier detailed view.
Home Page Widgets
Widgets can be added, positioned and configured by the administrator to present data from various sources like live feeds, promoted materials or other data sources:
- Static files (JSON format);
- RSS20/Atom feeds;
- Any source connection;
API Ready, Bridges, Back-End Server
Software can interact with Muse components by using various APIs. Federated Search Engines, Discovery services, Crawling, Harvest or ETL processes are usual examples of how our partners are using Muse.
Depending on the desired software integration, Muse Bridges implement the API over various protocols: Muse Web Bridge, Muse Z39.50 Bridge, Muse SOAP Bridge or other enterprise software specific implementations.
Muse can be transparently used as a back-end server for any portal or enterprise software to extend its searching or discovery capabilities.
Administration, Multiple Administrators
The easy management via browser based console and their easy usage via enterprise applications give the users and administrators a pleasant experience. By means of administrator consoles there is access to any settings of the system. Each administrator can have access to its own set of administering rights, which can vary from creating administrator accounts to only checking if a source is working.
The granularity of the administering rights allow an extended set of actions as filtering which applications to administer, which sources, which groups of sources, if another administrator users or applications can be created and so on.
Administration consoles are fully integrated with Muse Source Factory such that the evolution of source can be seen in any moment. An administrator can decide based on these information if to update or replace a source. There also are tools to completely configure any source or even to report that a source or a group or sources are not working anymore.
Pre-built Data Source Packages
The Muse Source Factory keeps the details, stores, manages, and distributes the thousands of Muse Smart Connectors to Muse installations worldwide. More than 6,500 different Data Sources are accessed by clients based on their own credentials and rights, using our Muse Smart Connectors. This collection is continuously updated and maintained to keep up with the latest native changes.
Multiple Sources, Various Protocols
The big number of connectors used for a large variety content and source types, make The Muse Platform a leader in the branch. It connects via tens of protocols (Atom, HTTP/HTML, HTTP/XML, JSON, NCIP, OAI-PMH, RSS1.0, RSS2.0, SIP2, SQL, SRU, SRW, Telnet, Z39.50) to thousands of Data Sources providing either free or premium content, no matter their record formats (XML, MARC, HTML, ISO2709, Fixed position, Fixed length, Lotus Notes, CSV, Office Docs).
Task Scheduler for Maintenance and User Tasks
Muse Control Center can be used to execute tasks at any given moment. The tasks lists, either default or custom, can be used for various purposes, starting from regularly updating the connectors to sending emails to users or even generate usage statistics of the system they run on.
Muse Control Center helps users run saved queries on various sources at any moment, without their intervention. The results extracted can be either sent by email or stored on disk and provided at a later time on demand.
Administration tasks can be also managed from Muse Control Center by specialized tasks. Such tasks do jobs, as checking the installed sources, get the latest versions from Muse Source Factory, check for new versions, making backups after updates also.
Tasks can be set to run periodically and send new results to user's inbox. Thus, any new result that appeared in one or many sources will be instantly spot.
End User Authentication
End user authentication is handled by single sign-on techniques. Muse controls end users’ access to resources using either User/Password files, client IP addresses, standard or custom authentication methods (LDAP, IMAP, SQL, SAML, LTI, HMAC, Referer URL). Users are thus granted access to the application area containing the authorized resources with just a single point of access.
Extended Logging Capabilities
Each Muse Product generates log files which are kept in special designated places. Sizing of logs dimension and rotation rules are very flexible, matching any usage of the servers. There are also available many logging levels.
The generated logs are of various types, either in well-know formats or proprietary. The proprietary ones are used to keep statistical information or debug, functional information and are analyzed by proprietary tools. The other logs, in the standard formats, can be also analyzed by our tools or others, like AWStats.
Sessions, Searches, Users Monitoring and Statistics
Muse offers high level real time monitoring and logging. Muse reports various types of statistical information through JMX and log files. The logged data can be used to obtain any kind of information, from the standard access logs as for any web server, to statistic logs which keep detailed data for all the Muse activity.
There are reports, either in graphical representation or tabular for most activities: successful logins, successful queries, unsupported queries, session life time, number of searches, number of sessions per period of time and so on.
The statistics exported through JMX can be read periodically using a JMX client program and they can be used for creating sheets or graphics of the Muse run-time activity.
Being implemented in Java, Muse can virtually run on any platform the Oracle Java Virtual Machine can. The Muse Setup is also multi-platform, providing both, OS dependent and independent installation packages.
Load Balancer Ready
Muse is load balancer ready. Using complex strategies, it can be places behind a software or hardware balancer to cover high traffic and usage requirements. There are available comprehensive manuals that treat any case of balancing, starting with scenarios where balancing should be applied and ending with very detailed settings that must be applied to each Muse component.
Partner Source Factory
This product extends the main Muse Source Factory into an architecture which allows local Muse Source Factory (Partner Source Factory) installations at Partners for various reasons, as: sources download into local Muse Installations of the customers, source development at partners’ side. This can be even seen as a distributed architecture.
There are tools that maintain synchronization between Muse Source Factory and Partner Source Factory installations. Also, any client having a Partner Source Factory install can use the specially designed tools to develop and maintain its own source connectors.
Integrated Development Environment, a suite for building connectors. The suite is completely integrated with Muse Source Factory to allow authorized staff to view development progress. Using the Muse Integrated Development Environment, building connectors becomes a production line activity, and maintenance takes its toll, as there are available syntax mappers, data conversion generators, overnight builds and automated testing. All these are much easier and more efficient, reducing the weeks needed for hand coded connectors to days.