The flood of data in companies today causes a lot of time spent searching for and discovering information. Transinsight’s Enterprise Search module solves that problem. With the help of semantic search engine technology, you can filter out any data you need from your local intranet or from the web! Enterprise Semantic Intelligence® provides a parallel search in databases, file systems, mail servers, and desktop PCs. Various data formats such as office documents, pictures, and tables can be browsed via an award-winning visualization.
Information management is absolutely important when it comes to the quick execution of processes. A lot of data comes from every department all over the company. This leads to complexity and, at worst, to a loss of information. This is caused not only by lack of organization but also because about 80 to 85% of enterprise data is unstructured. That information can’t be browsed with a normal keyword-based search engine. The effort of handling all that continuously growing data increases exponentially. Transinsight helps you to find your information and to gain new knowledge from your enterprise data.
Semantic indexing allows the next generation of search beyond keywords. By using knowledge in the form of semantic networks — also known as ontologies — our Enterprise Semantic Intelligence® Server brings two significant benefits to users: the completeness of results is guaranteed through the use of synonyms and ontonyms (the children concepts) in an ontology, and semantic correctness is guaranteed through word sense disambiguation.
State-of-the-art Text Mining
The Enterprise Semantic Intelligence® Server integrates the latest algorithms originally designed, developed, and optimized in the Bioinformatics domain. They were applied in the BioCreative competition in 2007 and won in the category of identifying genes and proteins in free text and linking these entities with the correct database entries. Additionally, the incomparable quality can be evaluated in GoPubMed, a platform for the elucidation of gene, protein, and disease interactions.
Multiple Data Sources
Search your intranet, your desktop, selected internet sites, MS SharePoint, and local databases. The Enterprise Semantic Intelligence® Server works with most file formats and APIs to databases. Highlighting is available for many PDF formats so that insight can be identified at a glance. This semantic search federation guarantees completeness of the search results.
Multilingual Semantic Support
The software can handle ontologies represented in multiple languages. The text mining will find relevant texts (in other languages) with semantic quality. This means that navigation and search can be carried out in the user’s native language, but non-native texts can be retrieved, too. The advantage of this is the small amount of returned texts, which later can be translated at a reasonable cost.
Data Security and Personalized Access
The platform provides an elaborate user rights management system that not only addresses data viewing but also the access control to data sources and ontology branches/leaves. The system allows the differentiation of read/write; it also supports groups and fully works with LDAP.
In the Enterprise version the number of documents that can be semantically indexed is unlimited1. The limits for the Professional version are set to one million, and to 100,000 in the Small version.
1 The maximum number of documents ever indexed was 100 million.
In the Enterprise version the number of locations where documents can be stored and indexed is unlimited. The limits for the Professional version are set to 10, and to 2 in the Small version.
Query the system without typing keywords. By just clicking on a concept or on multiple concepts, the platform expands the query using synonyms and ontonyms, the children of a concept. This allows surfing the search space and discovering aspects that have never been seen before. Faceted browsing rather than keyword search gives insights at a glance and avoids tedious trial and error for different keywords.
The system also allows the design of advanced queries based on the background knowledge. A semantic autocompletion box shows the most likely concepts for a certain context. This simple but powerful feature speeds query design while simultaneously providing insights.
Search results can be exported to different formats such as RDF, XML, or plain text and to library managers such as EndNote, BibTeX, Citavi, or Refworks. Many customized exports such as MS Excel are available on request. Tailor-made exports for business intelligence systems can be provided also.
A variety of deep analysis becomes possible through semantic indexing. The system provides tables for authors, dates, countries, and cities (if applicable). Semantic indexing is important because, for example, the correct disambiguation of author names – such as ’Miller’ – is necessary in order to achieve correct results.
This feature allows individual customization of the system. Users can pick concepts of the background knowledge of their interest. Favorites appear in the my search section and can be tailored and changed easily.
Integrated Semantic Community Platform
Connect with professionals who share similar interests or research activities. The integrated semantic mash-up identifies collaborating people on the internet or within your organization.
Users can easily cure the system in case of possible wrong classifications. A semantic curator tool allows users to quickly improve the text-mined data. Self-Learning Semantics let the system automatically learn over time. The accuracy of classification attains 95-100%. In contrast to other systems, this unique folksonomy feature constantly improves the engine without the unpopular process of tagging documents by users.
Queries as Favorites
Complex queries containing numerous concepts and keywords can be stored as personal favorites. Your stored queries can be quickly run whenever you want. This convenient feature allows you to precisely tailor semantic searches together with normal keywords not contained in the background knowledge.
The Discovery feature allows users to find concepts bridging two different and apparently unconnected concepts. This unparalleled feature, which becomes possible through semantics, allows true knowledge discovery.
Personalized Semantic Alerts
This feature allows making favorites and queries active. Users can customize the behavior of favorites and stay informed when new information becomes available. The alert is sent by email or can be observed within the my search section.
Integrated Semantic Blogging
This tool allows collaborative blogging with an ontology in the background. The system analyzes the current writing context and, via semantic autocompletion, suggests the concept rather than just words to enter. The web-based system enables the generation and editing of large documents online, and text mining is done on the fly.
Edit Knowledge Tree
A simple version for editing the knowledge base is provided by an in-place editing function in the tree. It does not provide the full flexibility of the ontology editor or creator, but you can customize concept synonyms and descriptions quickly and easily.
Customize your own ontology, your own “table of contents.” Easily include or exclude ontologies from the Enterprise Semantic Intelligence® Knowledge Server and Explorer. The system was developed in close collaboration with our customers and was designed with the goal of keeping it as simple as possible, but not simplistic. The software is designed for domain experts with little or no knowledge in ontology design and knowledge engineering. The system supports OWL and OBO as well as ontologies given in simple lists in text or XML formats.
Import/Export of Ontologies
In order to interoperate with other systems, this feature allows exchanging ontologies with other systems such as Protegé and OBO Edit. Other custom formats can be requested also.
This worldwide, unique tool allows the semi-automated creation of ontologies. The knowledge-base creation is set up in three simple steps: 1. Concept creation, 2. Definition suggestion, and 3. Ontology suggestion. The tool is made for domain experts with little or no knowledge about rather complex ontology editors such as Protegé. The software was used by one domain expert to create an ontology of about 20,000 terms having 50,000 synonyms in two months (www.Go3R.org – semantic search for alternative methods for animal trials).
The concept creator is part of the knowledge-base creator and allows creation of relevant concepts in the context of a certain topic. For example, for “aviation” the system would create the following list of related concepts: aviation, flight [GO:0060361], aircraft, pilot [IMR:0700908], air [281065, ENVO:00002005, PRO:000022063], spaceflight, safety. This tool helps to create an ontology when the knowledge-base creator is not available.
This tool easily generates definitions for newly created concepts. The concept itself is submitted as a query to the internet and to custom databases, and the returned results are screened from “is a” relations. The user compiles a selection from an automatically generated list and can then fine-tune the final definition with a small text-editing field.
Once a concept is created it has to find its place within an (in most cases) already existing ontology. This tool generates a list of most-likely positions the user can choose from. The concept of interest is then automatically incorporated within the ontology and can be used immediately for semantic search.
|Enterprise Semantic Intelligence® Server|
|Improved Text Mining|
|Multiple Data Sources|
|Multilingual Semantic Support|
|Data Security and Personalized Access|
|Enterprise Semantic Intelligence® Explorer|
|Integrated Semantic Community Platform|
|Self Learning Semantics|
|Queries as Favorites|
|Personalized Semantic Alerts|
|Integrated semantic blogging|
|Enterprise Semantic Intelligence® Studio|
|Edit Knowledge Tree|
|Import/Export of Ontologies|
“GoPubMed is ingenious”
“Scientists need a solution to literature navigation not a learning curve. GoPubMed makes it simpler to find papers that I want to read.”
”GoPubMed from Transinsight has inspired us because its interface really stands out from the multitude of scientific information systems. For us it’s the perfect example of combining complex searches with a first class, easy-to-use interface."
"This is one of the best ways I’ve seen to wade through the sometimes massive results of a PubMed search."
”I love this search engine. Iteration is simple, effective and dynamic. From the 3 million plus authors and the 15 million plus articles within PubMed this engine gives the most time-effective result return"
”Well done on GoPubMed. It has brought the PubMed interface into the 21st century! ”
"20.000 visiteurs uniques par jour bénéficient du projet de navigateur intelligent..."
“GoPubMed is mostly self-explanatory: I suggest people just wander over and try it out.”
“Searching in GoPubMed provides more options for sorting to the searcher than PubMed.”
"Transinsight wins German Federal Ministry of Economics and Technology award"
"Gopubmed contains a huge knowledge base of information and top terms. This search engine won a 2009 design award for the “best of the best” semantic filters."
"GeneCloud developer Transinsight was also awarded the German Innovation Prize IT for its “outstanding developments in Enterprise Semantic Intelligence at CeBIT 2011."
“Another tool I found fascinating to try out is GoPubMed which identifies leaders in a field. In GoPubMed insert your favorite topic in the “Find” box on the homepage and, once you have those results click on “Statistics” to see a summary description of topic, top authors, institutions, journals, and in vogue years for that topic.”