The Knowledge Discovery Platform RetrievalWare 8.1
As the search market has heated up, Convera has been positioning itself as a lean player in this space by aligning resources to two primary goals: consolidating and furthering its extensive lead in the government information retrieval market; and gearing up for an offering that confirms its position as the most complete provider of enterprise-class discovery services.
This is being achieved through the company's latest offering, RetrievalWare 8.1. RetrievalWare is a complete knowledge discovery platform for information-intensive enterprises. RetrievalWare reduces the time and cost of finding information in high volumes of structured and unstructured data. This improves business process efficiency and increases return on information assets. Convera's RetrievalWare can access information across all types of portals, applications, repositories and file systems. It can use a variety of indexing, concept and entity extraction and content filtering methods regardless of content type—unstructured, semi-structured and structured. RetrievalWare's unique Knowledge Discovery Services—such as profiling, concept search, persistent and dynamic classification, personalization and collaboration—provide context-relevant, precise and unified information retrieval. RetrievalWare also provides comprehensive multilingual and domain-specific knowledge structures including semantic networks, thesauri, taxonomies, integrated KM applications and multimedia asset management applications.
RetrievalWare Web Services
Convera has made it even simpler to embed high-quality search and classification capabilities into a number of enterprise applications. The new RetrievalWare Knowledge Discovery Services, based on industry standard Web Services and XML, dramatically reduce the time, cost and risk of enterprise application and portal integration by mixing and matching large, re-usable software components and bolting together complex high-quality knowledge discovery engines—all working across multiple applications. Because the same RetrievalWare software services are also available as J2EE-based Java Services, it is possible for the enterprises to architect RetrievalWare-based solutions across multiple operating systems and platforms—.NET or J2EE. Reduced integration, customization and localization time reduces start-up cost. Ability to scale with growing content lowers cost of ownership. Standards-based, open architecture, cross-platform support and secure access to information reduce risk.
Continued Focus on Search Needs through Industry-Leading Relevancy Models
Optimized Search Precision & Recall. Through the combination of keyword indexing, advanced linguistics processing and a variety of basic and advanced search methods, RetrievalWare achieves optimized precision and recall across vast and diverse information sources. Boolean, pattern and concept modes can be used independently or interactively. RetrievalWare's intelligent semantic indexing utilizes general-purpose and domain-specific semantic networks in order to complete semantic analysis, and discover synonyms, broader and narrower terms and related terms. This improves search results by identifying all relevant documents to a search rather than just retrieving document that match keywords.
Discovery of Critically Relevant Information Using Taxonomies. Categorization and dynamic classification represent a behavioral and technological leap for users and organizations alike. Rather than being forced to fit searches within the constraints of inflexible categories, users can dynamically create their own information categories based on the context of their search at the moment. Further, those categories can interrelate and display information from widely disparate sources and locations, permitting users to discover knowledge that might have otherwise remained hidden.
Bringing Context to Content with Taxonomies and Classifications. RetrievalWare's advanced categorization and classification solution enables organizations to discover links, knowledge and expertise embedded within otherwise diverse and scattered information sources. Unlike other categorization and classification solutions, RetrievalWare uses scalable and consistent taxonomies for categorization and flexible and pragmatic classifications for information access.
Categorization with Taxonomies. RetrievalWare's categorizer automatically extracts concepts from documents using taxonomies and creates a semantic signature (metadata) for each document. Taxonomies contain thousands of concepts organized in consistent hierarchies with generic-to-specific relationships.
Automated Classification for Information Access. RetrievalWare's Classifier automatically classifies documents using concept rules and ranks defined in classifications and matches documents to relevant categories.
Dynamic Classification.Users can launch a search and automatically classify the results based on pre-defined or dynamically generated classifications. The underlying taxonomies can consist of Convera's pre-packaged industry taxonomies, customer-defined taxonomies or custom taxonomies supplied by Convera's Taxonomy Development Partners.
Entity Extraction: Identification of Named Entities, Patterns and Relationships
Enterprise data sources contain critical references to organizations, people and products within text fields in databases, e-mails and portal pages. Information in documents containing these entities is invaluable when it is related to information about the same or related entities in other documents. As the volume and frequency of enterprise content grows, it becomes very expensive and time consuming for knowledge workers to identify the patterns and relationships critical to discovery. RetrievalWare's Entity Extractor solves this problem by extracting hidden information from real-time and archived data sources and by providing relevant access to information which enables knowledge discovery. RetrievalWare's entity extraction is similar to categorization in the sense that it automatically extracts tags from content. While the Categorizer uses taxonomies to extracts concepts, Entity Extractor uses known entity lists or pattern-matching algorithms to extract entities from documents. Users can launch a search and automatically filter the results based on the categories of entities. The underlying entity cartridges can consist of Convera's pre-packaged lists of known entities and rules. In addition, subject-matter experts and librarians can import their own lists and customize their own rules using the RetrievalWare Knowledge Workbench.
Real-time Monitoring of Information and Alerting
Profiling. Profiling is a content-filtering feature that enables real-time monitoring of information and real-time execution of queries on individual documents in live data sources. When a new document matches a profile, the system immediately adds that document into the user's personal folder for that profile, and increments the counter for that folder. When the user clicks on that folder, the system displays the latest matching document(s) at the top of the results list and enables the user to find the "hit" or "hits" which caused the document to match the profile. Queries used in profiles offer an ability to track personalized information in real-time via various means which eliminates the risk of missing critical information or delayed access to relevant information.
Alerting. An e-mail alerts feature, in conjunction with profiling, enables users to keep track of late-breaking news, constantly changing Websites, portals, content management systems and collaboration systems automatically and in real-time. Using a custom client application for e-mail alerts, users can specify how often they wish to receive alerts for a given profile. For example, a user may choose to receive e-mail alerts on news items on an hourly basis as opposed to e-mail alerts from a CRM system on a daily basis. Whenever a story on a specific topic breaks, or when a subject of interest comes up in a data source, users can be alerted via e-mail.
On a Continuous Path of Innovation
Convera remains committed to offering leading-edge discovery services and adding to these over the next few years. Already the information retrieval landscape has been permanently changed due to Convera's unique approach to categorization and dynamic classification. The addition of knowledge structures such as semantic networks, taxonomies, thesauri and entity lists that can be mixed and matched to best achieve effective location of relevant data is also in lock-step with increasing user awareness of the richer content that can be exposed intelligently. As the semantic organization of content takes on more and more importance, Convera is very well positioned to continue to lead with best-of-breed capabilities for increasing the findability of critical information nuggets needed in the daily tasks of knowledge workers and analysts.
CONVERA 1921 Gallows Road, Suite 200 Vienna VA 22182 PH: 800 755 7005 or 703.761.3700 FAX: 703.761.1990 E-mail: info@convera.com Web: www.convera.com