[j]search is a semantic search engine and DRM for the Single Point of Entry to everybody's personal Knowledge-Space in Dental Medicine. [j]search searches the web for content, especially scientific content, day by day and has interfaces to paid scientific content of several publishers. It finds relations between knowledge objects and combines them to an individual knowledge-space, based on the demand of every user. Content sources of [j]search are sorted in realms, that can be composed depending on the demant of the users of a website. Together with the “Relevancer” engine, a personalized profile for the user is generated. [j]search comes along with a powerful DRM that provides information about prices of paid content and performs a comparison with the individual subscriptions and rights of every user.
- interfaces to paid content of several publishers
- interface to open source databases
- integratable in websites and platforms
- powerful backend for DRM
- admin friendly UI for content curation
- user friendly UI for faceted search
- variable query types
- ranked searching - best results return first
Integration in third-party applications
In addition to existing plug-and-play components for Liferay, which perform a wide variety of tasks as complete applications, customer specific applications can be developed with the help of [j]search on the basis of powerful Java or Python based interfaces. [j]search provides several high-level integration libraries (full featured Java and Python APIs) and a powerful REST API that allows a caller to extract features from passed content.
JSON, REST and SOAP
[j]search can be integrated in almost every website or platform via JSON, REST and SOAP interfaces. [j]search supports a virtually infinite type of custom metadata fields. [j]search has a broad range of metadata attributes, which enables the user to narrow the results according to its needs, e.g. based on attributes like the document type.
Using the [j]search search engine, several functionalities have been developed that work in the open source portal liferay, one of the most used portal technologies worldwide. For liferay portal, two [j]search portlets are available with the following functions and features. Similar functions and portlets could be integrated also in other systems.
Delivers information about similar documents. If termVectors are not stored, MoreLikeThese will generate terms from stored fields for example from Simple Anno. This method returns similar documents for each document in the response set.
[j]search supports enhanced author support along multiple author databases. Information about authors will be aggregated and stored by [j]search. Currently there are more than 1.2 million scientific author and expert records available.
[j]search - index can be split into multiple shards.
[j]search can query and merge results across those shards. This kind of decomposition of an index allows distributed storage on different servers and server locations. Indexes can be hosted on data storage in different places. Unlike traditional centralized search engines, work such as crawling, data mining, indexing, and query processing is distributed among several peers in a decentralized manner where there is no single point of control and no single point of failure.