Our clustering engine is built upon our proprietary technology which analyzes and processes the raw content stream and smartly generates the cluster hierarchy. The core clustering engine in itself is only responsible for generating a smart and meaningful cluster hierarchy using a proprietary clustering technology know-how developed by us. Our clustering algoritm consists of a complex set of data and pattern analysis, cognitive analysis, optimization and normalization processes followed by a sophisticated inclusion-exclusion phase. As a result, the cluster hierarchy and structure, although machine generated, reflects a human-like behavior rather than a machine or methametical process.

Instead of just producing a large number of result pages a clustering search engine groups (categorizes) the results into conceptually meaningful hierarchy of clusters otherwise known as folders. No prior categorization or cluster (folder) definition is required; everything is done on the fly and without any human intervention. Our innovative auto-categorization algorithm takes care of it all. Hence all the results are presented as a hierarchy of meaningful clusters (folders) they are conveniently accessible by the searcher in a much easier and faster way.

Another important aspect of our clustering engine is, that its behavior and clustering criteria can be changed and reconfigured to create to create a special purpose clustering search engine or information service. As a result we can build a clustering search engine for general purpose search (like web search) or a highly targeted specilaity search engine for searching a technical subject speciality or financial information or something similar.

The ability of the clustering engine can be further expanded by adding the optional Facetted Clustering Module (often referred to as Multi-threaded Clustering). Facetted Clustering is useful to generate the cluster hierarchy based on additional cluster (folder) definition to create cluster grouping with strict relevance and/or relation to a specific topic or criteria. This module is deployed in special cases, as for example in a centralized or global information search portal. Another example of the facetted clustering module is shopping search where products must be organized and displayed strictly by product type or category or outlet.

The core clustering engine usually comes embedded in a suitbale version of EMS/Integra search frontend to create a fully operational information service or portal.

Advantages:

Dramatically shortens 'Seek-and-Find' time.

Easy to understand search results. Any one can easily locate sought information without any in depth knowledge of search engines.

Organization and cluster hierarchy is aligned to natural human behavior and instinct. As a result, it is user friendly, well percived and easy to use.

Perfect search and analyzing tool for researchers, librarians, educators, journalists and knowledge workers.

Results always appear to be with-in easy reach. Searchers do not have to browse through pages after pages thus losing track of previously seen listings.

All this together contributes to better user experience and increases service popularity and effectiveness.

More Advantages for commercial portals and search engines:

Ability to show highly targeted paid listings, PPC/partner feeds, advertisements in line with the cluster hierarchy concept. As a result shown listings are by far more relevant to the currently viewed cluster/folder.

When deployed in a open access web portal or search engine, clustering search tend to significantly increase page views and valid click-thoughs.

Due to the fact, that results are categorized with in conceptually meaningful folders searchers find them more relevant and reliable.

Better search experience encourages visitors to remain loyal.

These are just some of the major features. For further details and to find out how our solutions can help achieve your aim in a cost effective way contact our sales team.

Back - Jump to Top