High Performance Fuzzy Data Matching and Search for Large Databases

Is Your Search Engine Really Optimal?

An effective search engine must be user-friendly. However, that is a more complex objective to reach than meets the eye. Such a search engine is not a replica of popular web search technologies made to search billion of web pages. It's not sufficient either to have best crawling and indexing system capable of reaching all your data and documents wherever they may be on the network. It's far more. It needs to take in account the particular nature of your data and it needs to offer the best possible search matching technology available for each data type at all levels (indexing, matching, and displaying).

Experience shows that search usage may increase by as much as 5 to 12% by simply introducing approximate matching features to traditional search engine applications. Popular web search applications have made users more savvy about using search engines. However, their expectations have also risen as a result of the ubiquity of search. An application, a web site, a public service, a directory, or a data bank that makes it easy to find its content and data attracts and retains more users. That’s just how simple the strategy is. That’s always good for business and customer satisfaction. When it comes to the tactics, solutions can be complex. So come to the specialists with the best tool in the industry: Exorbyte and MatchMaker.

How User-Friendly is Your Search Engine?

Read how a specialized structured data search platform like MatchMaker can bring ground-breaking benefits to all your search applications.

Get easy instant access to this white paper!

People only return to web sites when they have established that they can find what they need time and time again. To many web site and application managers, this translates in fast exact matching and deep indexing of all their content. That's just not sufficient. There are many reasons why search queries result is zero results or poor results being returned to users:

misspellings or typos in search query (15-25% of all web queries are misspelled)
misspelling or typos in indexed content
alternate spellings in indexed content or search query
close-matching content indexed not recognized by matching engine
related content indexed not recognized by matching engine
synonyms, acronyms, and other related terms not matched
poor use or no use of taxonomies and categorical structures in the search index
inappropriate use of approximate matching algorithms (n-gram etc.), phonetics (soundex) and semantics to detect results
spellchecking technology that only corrects most common mistakes (stemming, plurals, etc.)
slow response time of search engine
inappropriate choice of advanced query parameters by user
inappropriate ranking of search results
search engine doesn't suggest relevant alternate queries
search query suggestions result in no results
etc.

As you can see above, the failure can reside with the content searched, the search engine, its configuration, the user query, the quality or speed of the results, and many other variables. At Exorbyte, we never blame the user, or our customer, or our customer's content. We just solve these problems so you or the user doesn't have to. We have built MatchMaker Search to be the ultimate facility to remedy all these known scenarios. MatchMaker search is built to add to or replace your existing search engine seamlessly. MatchMaker Search gives you the control you need to resolve the most complex search challenges.

Does your search engine give full access to your data or content?

MatchMaker Search brings unprecedented value to enterprise search, Web portals, directories, and paid search solutions.

Benefits

Satisfied users trust you and come back

A powerful search solution like Exorbyte’s MatchMaker widens the narrow search corridor of exact search. The user receives helpful results despite misspellings, alternate spellings, etc.

Ultra-fast fuzzy data matching

MatchMaker is ultra-fast: even with millions of database records, the system responds within 10 milliseconds, resulting in highly satisfied users. In addition the results ranking is completely transparent, which is important for optimizing performance.

Database Independence - Fast Implementation

MatchMaker runs in parallel to existing data infrastructures. This allows the usage of low-cost hardware, requires minimal changes to the existing IT infrastructure and provides database independence. The implementation is fast and simple and can be completed within days. The architecture guarantees full scalability and usually reduces the database load.

Synonym dictionaries and taxonomies

The search performance of MatchMaker can be further enhanced by the addition of alias dictionaries, taxonomies, and various other semantic features. Exorbyte offers a powerful software tool to develop alias lists from query logs. MatchMaker also supports taxonomies and categorical structures in its indexing and matching technology. This tremendously enhances the ability to locate and suggest relevant results.

System Performance and Requirements

Results returned in less than 10 milliseconds even with tens of millions of records in the index.
OS: Windows, Linux
APIs: Sockets, C++, COM, PHP, Java, Tcl, Python
RAM: Approx. 80 MB per 1 million records