Application Scenarios & Use Cases
On top of the PoolParty service-interface a variety of semantic web applications can be realized, for example:
-
Corporate Thesaurus & controlled vocabularies
Any form of controlled vocabulary, be it a taxonomy, thesaurus or ontology greatly helps you in managing, organizing and finding your content. We find that thesauri have the best cost/benefit ratio of all controlled vocabularies, as they are easy to create and to understand, while providing all the semantics that are needed for a wide range of applications.
Terms in a thesaurus are not mere keywords, but controlled concepts that have serveral so called labels attached. The labels are all the words (such as synonyms, abbreviations, misspellings or translations in other languages) that can serve as a name for a concept. E.g. a concept might have the labels United Kingdom, UK, Great Britain, and Großbritannien. All concepts are connected in a hierarchy through statements specifying “broader” and “narrower” relationships, e.g. European Union is a broader concept of United Kingdom which in turn is a broader concept of England. There is also the possibility to connect terms in a non-hierarchic way, by stating two concepts are in some way “related” to each other, e.g. concepts Shakespeare and United Kingdom are related.
Thesauri are of great value when it comes to integrating data within your organization as well as with external sources. All the use cases below can be realized with a thesaurus and the services provided by PoolParty. -
Tag Recommender Systems
Tagging can be very helpful in increasing the findability of information, but we find that in convential tagging systems many users rarely apply tags and that the inevitable inconsistencies in tagging behavior considerably hold back its usefulness.
Having a thesaurus backed tagging system allows for consistent use of tags, as tags are no longer mere words, but concepts that include synonyms, abbreviations, misspellings, etc. Your content can be automatically analysed by PoolParty’s Natural Language Processing modules and related concepts can be suggested as tags. The user can choose these controlled tags, or start entering their own tags, which prompt the system to suggest concepts from the thesaurus that match the entered string (see use case Autocomplete). In this way PoolParty lowers the barriers to tagging and improves social tagging & social bookmarking in your systems. -
Autocomplete
An autocomplete functionality comes in handy for supporting the user in both searching and tagging. Once a user starts entering a keyword the system compares the entered string to the terms from the thesaurus with all their synonyms and suggests possible completions. Those of course are concepts, not keywords: When someone enters e.g. “jav” in a search or tagging widget, the system can suggest “Java (programming language)” and “Java (island)” to the user, thereby providing concept disambiguation and tagging consistency.
-
Semantic Search Engines
Being able to search for concepts instead of keywords not only increases precision and recall but also allows for new ways of user friendly search mechanisms. No matter whether users searches for Great Britain, UK or Großbritannien, they will get the same results with the documents containing any of these words. The system can offer a moderated search, where terms related to the search terms are suggested to the user, so she can include or exclude e.g. documents refering to London or Shakespeare. An example application of a semantic search engine can be seen at ecoi.net.
PoolParty is making innovative use of Linked Data. When creating a thesaurus with PoolParty you have the possiblity to link concepts to their counterparts on external systems. This means a thesaurus manager can retrieve a globally recognized identifier for e.g. the concept London and get additional information like abstracts, category information, related concepts, longitude, latitude and pictures of London. In this way one can enrich thesaurus concepts and consequently the semantic index in which document tags and other meta data is stored, thereby improving search results and content similarity calculations. -
Faceted Browsing and faceted search
Faceted Browsing and Faceted Search are a new approach to help users explore available content in a structured manner. Whereas “advanced modes” of search engines are used only by a few power users, a user interface with faceted search capabilities can stimulate more users to filter and refine search results. It can be used to browse content according to the categories defined in a thesaurus or to further filter search results to include only content related to certain facets. An example application of faceted search can be seen at emteba.at
-
Similar Documents Recommendations
In the same way as with Semantic Search a concept’s labels and relationships can be used to improve similarity recommendations. When a user looks at a document the system can use its semantic fingerprint derived from the concepts it is about and the information about their related concepts to suggest similar content. From this follows that a document mentioning London will have some similarity to a document mentioning the United Kingom if the relationship between these concepts is modelled in a thesaurus. PoolParty’s use of Linked Data (see Semantic Search) is paramount in improving the semantic fingerprint of documents.
-
Semantic Wikis
Your wiki can greatly benefit from thesaurus backed autocomplete, tagging and search services provided by PoolParty. Additionally semantic wikis can provide the users with forms for entering data and the fields of the forms can be can be automatically marked up with predefined semantics from a thesaurus or ontology. PoolParty makes use of semantic wiki interfaces to provide an end-user interface to maintain thesauri in collaborative working environments. See an example here.
-
Personal Information Systems
Using a personal thesaurus greatly enhances the findability of your information. Modelling a thesaurus according to your ideas and needs also helps you to acquire deeper understanding of a knowledge domain you are interested in. PoolParty can be a great tool for e.g. information workers, teachers and students for gaining and organizing knowledge.