FAQ's
This page will answer most frequently asked questions about the Visvo web crawler.
- Why is Visvo special? How is it unique?
- Is Visvo a metasearch engine that uses results from other search engines to provide searches for its users?
- What's wrong with search now? Why launch a new search engine?
- How does Visvo benefit the user experience in search?
- How does Visvo benefit advertisers?
- Is Visvo always available 24x7?
- Does Visvo support usage by enterprises?
- How does Visvo compare to NLP based search engines that have been launched or will be in the future?
- How does Visvo plan on dealing with the strong competition in the search business?
- What other services are planned by Visvo?
- Does Visvo plan to launch in other countries and other languages?
Visvo divides web content into major categories and subcategories that return highly focused results, eliminating irrelevant information. The Company's search algorithm is significantly better than PageRank algorithms and specialized vertical search algorithms because it provides:
- Machine learning algorithms that can understand the context or meaning of the web pages they process;
- Automatic classification algorithms that can correctly categorize content according to its context and meaning;
- Integration of categories into a search engine to create a one-click search model;
- "One-click" drilldown into areas of interest, eliminating tedious multi-word entry, and perfectly adapted for cell phone use;
- Discovery and entertainment. Distribution of search terms by categories encourages users to explore and surprise themselves by following their search terms into new areas.
Visvo does not rely on link analysis as its major component. Unique among its competitors, Visvo uses page content analysis as the major component in its relevancy score. Visvo's algorithm works by breaking up pages into a series of noun phrases or "tokens" that represent the type of information that exists in each page. These tokens are then used to automatically classify content into its most probable category by meaning and relevance; tokens on a particular page are matched to similar tokens in predefined categories. The algorithm reads, learns and weights actual page content. Visvo uses commodity hardware, and proprietary improvements to open source software, to create inexpensive yet powerful parallel and distributed computing architecture and virtually unlimited processing power and storage space. The Company's proprietary word/phrase/page aggregation and segmentation algorithms are so robust that Visvo's classification will result in search results that match user intent.
- Irrelevant results and impatient users
- Only 17% of users find what they are looking for
- Frustration from typing in more words or looking at more results pages
- Result: 66% of searchers visit at least two search engines
The opportunity to launch is there if the search engine can provide:
- Significantly better, targeted search
- Better results for users
- Better-qualified customers for advertisers
- Engaging discovery and exploration potential
- Convenience of a "one-click" drilldown model
Harris Interactive reports that about 80% of internet traffic begins at a search engine. Of all Americans that use the internet, about 85% use search engines, or more than 107 million people in the United States alone. Search marketing is expected to generate over $7.0 Billion in revenue in 2005, growing to nearly $24 Billion by 2011. Visvo's opportunity lies in capturing a reasonable share of this very large and very high growth market through offering a differentiated search product with high utility, high authority, high confidence, and ease of use.
Since the early days of search, search companies have understood that delivering personal search refinement is critical to user satisfaction. In the "attention" economy, information is plentiful and human attention is scarce. Huberman's (2001) survey indicates that 0.1% of Websites capture 32.3% of activity, indicating that the vast majority of Web content languishes in obscurity. A recent Comverse survey found that only 43% of professionals find what they want after several search attempts. Search engines often do not know what the user wants based on the search query. Whether by improved semantics, web "concepts", word clustering, or specialized verticals (such as Travelocity), search companies know that increased user satisfaction requires improved search structures that better understand user intent not merely user word entry. Refinement requires improved classification and clarification of user intent. Visvo believes that its search engine is the only engine that satisfies this requirement.
It is Visvo's view that, as the search market matures; web-based advertisers will increasingly acquire their clickthroughs through improved context-sensitive advertising that delivers a more qualified user to the advertiser's message. Specialized and targeted categories of interest that both create and refine their own user base ensure improved context and value to advertisers.
- Highly Differentiated, eliminating ambiguity. Visvo focuses on determining, classifying and returning what the user actually means in his search, not merely what word the user happened to choose to enter into the search box.
- Automatic. Visvos engine automatically categorizes content into a hierarchy of categories, providing the relevancy of a directory with the speed and depth of a search engine.
- Highly Efficient. The ability of Visvos engine to search the web and create and manage highly specific categories automatically without user intervention is a significant step forward in extracting the most useful responses to user queries.
- Bounded. The Visvo algorithm ensures that additional "drill down" searches in a category are conducted only within the bounds of that category. "Page Rank" and clustering algorithms are unbounded, and merely serve the users query (even with more specific words) against the entire internet.