Today hakia added a hakia highlighter to their “meaning-based” search engine, producing a highlighted sentence inside a search result. The bigger announcement is tomorrow, when hakia will launch a scoop button - a browser plug-in that not only highlights text, but when you click on a result page it scrolls automatically to the highlighted passage, enables you to save data to your computer, and more customization features that we'll discuss below.
Both of these new tools allow for faster more relevant result selection and additional utility for users.
hakia highlighter
The hakia highlighter upgrade addresses what Melek Pulatkonak, COO of hakia, termed "Click-Thru-Itis" - or clicking through a link to determine relevance. Traditional search engines often force users to click links excessively in order to determine (based on limited information) which result is relevant. The hakia highlighter alleviates the need for excessive clicks by displaying meaning-based, uninterrupted sentences in the search result. Determining relevance and providing an educated selection of URLs is improved because of the added content and context provided by these sentences. Broken keywords and phrases simply do not provide enough information.
Let's examine what typically happens when we do a search on Google. For the search query: "What does it mean to cross the Rubicon?" the Google results are almost always more difficult to "filter" because of fragmented meaning. The real problem is in the disparity between relative results and how people are forced into making decisions based on this broken information. The Google results to this query vary from the rule of habeas corpus to a metaphysical discourse, but the bold keywords do not signal this wide disparity.
Note: fragmented keywords and phrases
Invariably, selecting from broken sentences will lead to unwanted visits to those URLs. Let's now look at what is revealed with the same search on hakia. The highlighted sentences provide more information and relevance for deciding. Results that display no sentences on hakia are ruled out all together and more obvious examples become readily apparent. In this example, extra clicks aren't necessary in order to significantly "narrow" a subject.
Note: I followed the highlighted sentences as much as I did the links
The hakia highlighter examples demonstrate the engine's ability to "think" semantically and display the process in a way that narrows selection options.
Scoop Bar
The scoop bar is a browser plugin that lets users apply semantic results in a unique way. Currently it is only available on Windows, but Mac support is coming.
When installed, clicking the scoop browser icon reveals a pull-down with several options. The home option refers back to hakia, where a search for "What is the specific gravity of lead?" renders a highlighted result as before; but clicking on a desired result reveals the page AND scrolls the page to the highlighted passage as below.
Note: The scroll bar is about 1/3 down a very long page on ballistics
The button in front of the highlighted passage has several functions. Clicking the pull-down arrow allows the user to highlight desired text and save the link and text to a custom folder. Alternatively, clicking the "scoop and save" icon saves the link and text to a default folder or file.
Note: Additional highlighted paragraph and saving to a created folder
Additional buttons and functions allow for more customization and navigation. The options button at top center of the tool bar brings up the scoop and save functions, while the "my results" pull-down lists saved results and navigates to them as illustrated below.
My result function and destination folder
Perhaps a real world example is in order. If I were doing a post on ballistics, Hakia (even in beta) has provided me with faster and more relevant results, a method for saving/customizing results, and a point for the study of all the data gleaned. A user could perform the same function in any browser, but it is obvious that time and function would be lost comparatively.
Conclusion
Hakia is progressing to the point of expressing results in true "natural language" terms. Make no mistake, the comparative results of Google, Yahoo and hakia are not conclusively differentiated yet, but the progress of hakia is fairly clear in these examples. The way the "scoop" function directs the user to relevant links and then "scrolls" to the pertinent passage is evidence of the semantic engine at work locating relevant data. The implications of this are powerful and exciting for hakia and the rest of us. There is a long road ahead for hakia, but predicting outcomes is so often a function of watching the little things.
No comments:
Post a Comment