Alternative Engines
Semantic Search | Alternative Display | Data
Google, Yahoo, and Bing are seen as the traditional or (even) orthodox search engines - where the searcher types two or three keywords, the engine matches on these words (with some enhancements) and displays the results. In contrast, Ask has been somewhat alternative as of late because of its work with semantic technologies to parse the user's query. There are many others that employ other matching and ranking routines or approaches to display. Occasionally, some of these develop into first rate tools. Sometimes there is something completely different. Wolfram|Alpha is such a case: it is a "computational knowledge" engine that returns data as the answer.
Semantic Search
Many new engines are being developed that employ linguisitic analysis and semantic technologies to "understand" the question and the content - these are engines that do natural language processing. [Also see Search Syntax > Natural Language]
Hakia looks for meaning through semantic connections of words to concepts rather than relying on the standard keyword match. It attempts to understand the intent of the searcher's question through these connections and to match that to content that has been similarly analyzed. Hakia also works to assess the quality of the sources.
Hakia is excellent for short, discovery type questions for people, places, and some health conditions. It returns categorized results called galleries as a quick guide to the topic - such as this on on Ottawa.
You can easily compare Hakia's results to Google, Yahoo, or MSN (Bing) at Hakia's challenge site. Such a bold offering shows how much confidence Hakia has in the relevancy of its results.
Hakia has developed a browser toolbar - Hakia ScoopBar - that will automatically find the answer to your question on the selected web page. You can also use this to copy and save information.
CEO of Hakia, Riza Berkan, argues the case for semantic search in The Search for Quality on the Web in AltSearchEngines (Feb 2009)
"The underlying idea behind semantic technology is to teach computers how the world operates. For example, when a computer encounters the word “bill,” it would know that “bill” has 15 different meanings in English. When the computer encounters the phrase “killed the bill,” it would deduce that “bill” can only be a proposed law submitted to a legislature, and that “kill” could mean only “stop.”"
Factbites answers in sentences. It prefers factual questions - or just a word or two. There are fewer results but the answers are better because the sources used are encyclopedias and higher content sites. It describes itself as a hybrid of a search engine and encyclopedia. It won't have all the answers, but on many topics it can be a good start. (Made in Australia)
See this profile on Tecumseh, the Shawnee Indian Chief. Compare that to the more specific chief tecumseh was buried.
Kosmix.com Added Dec 2009 WSG Pick
Kosmix excels at creating a portfolio on a topic drawn from the range of web resources: reference resources, images, video, web search, blogs, tweets. Its algorithms employ a taxonomy of categories to select relevant and appropriate sources for that topic. You'll see it in the page on arctic exploration, which in addition to the variety of results, lists several related categories as a discovery aid.
Follow the Kosmix blog for new developments in search.
alternative Display
Other alternatives are using new visual displays or ways for the searcher to view and work with the results.
DuckDuckGo.com Added Dec 2009, WSG Pick
DuckDuckGo cuts the clutter - it is effective in blocking spam in its crawl of the web, and it facilitates discovery on general topics.
The front search page has options to search the web, search information sites (no consumer), shopping sites, or to "Duck it" to get topic summaries. Topical analysis is the key attraction to this search engine.
- DuckDuckGo recognizes words that have multiple meanings and will present those as choices. Eg bonds.
- On a broad topic, the web search will give a capsule description and show related topics. Eg arctic exploration.
- The Duck It choice will show best pick and Topic Summaries taken from top ranked results. Eg conservation of boreal forests
Creator Gabriel Weinberg talks about his search engine in this under-a-minute
"elevator
pitch" (Nov 2008)
Quintura (www.quintura.com)
This visual search engine uses the Yahoo database for web searching. It clusters results in a tag cloud which can be manipulated to alter the search. Quintura will also search Yahoo images and video, and Amazon books.
TouchGraph touchgraph.com/TGGoogleBrowser.html
TouchGraph is also a data-visualization tool. It uses Java to display a map of nodes with results clustered around. Colour coding matches the network to the list of sites in a side panel. It can be used to search Google, Amazon, or Facebook.
Viewzi has many forms of presenting results from Google, Yahoo, Flickr (photos), audio sources, videos, blogs and more. Sample them all. For a search on a topic, such as information management, try the Powergrid (4 X 2 grid of thumbnails from Google and Yahoo) and the Web Screenshots for paging through the results one by one.
Data Search
Wolfram Alpha Added Dec 2009
Wolfram|Alpha's "goal is to make all systematic knowledge immediately computable and accessible." As a "computational knowledge" engine it thrives on data - it collects it, massages it, and delivers it as answers. It does quite well on questions concerning company financials, market data, population, weather, definitions, nutrition and calories, and it can be used for mathematical functions. Scientist and mathematicians will love it. This is not a place to ask word questions. Best approach is to explore the page of examples to learn whether it has answers for questions you need to ask.
Where to Next?
Searching for multimedia.