CHiPS is an ontology-based semantic search engine for an enterprise document repository. Unlike traditional search engines that find documents by matching literal keywords given in a brief search string provided by the user, semantic search works by finding documents that are semantically similar (i.e. similar in meaning) to the search query.
The key advantages comparing to traditional search systems are:
Instead of a sparse handful of keywords, you can make use of full sentences and paragraphs that describe what you are looking for. The search query in CHiPS can be a single question or an even an entire document which enables more applications: search for similar documents, finding relevant forum branches for a verbose new message, etc.
Discover new directions of research through the Related Concepts cloud comprised of ontology concepts found in the result set, but not present in the query.
Each search is cached and assigned a unique bookmark link in order to enable collaborative data analysis between teams. Just send a collaborator the search bookmark link, and they will see the same search and results.
The CHiPS client side requires only a web browser to access. Supported browsers include Firefox and Chrome.
CHiPS web server requires Java 1.6 or later and a connection to a Postgres 8.3 or later database. The web server operating system can be either Windows or Linux.
CHiPS uses Lymba's K-Extractor software to perform natural language processing. K-Extractor is available for x86-based processors, 2GB or more of RAM are required. It has been certified on RedHat Enterprise Linux 3 and RedHat 9, CentOS 5, and Oracle Enterprise Linux.
The CHiPS administrative tool requires Java 1.6 or later and at least 1GB of RAM.