Business

Blinkx: Desktop, web search tool

THIS EDITION of NetSpeak brings to your attention the new web/desktop search tool, blinkx, which has already garnered a lot of attention from Net technology observers.

In response to your search query, a normal search engine brings you the data sources available on the Net with the required content.

However, apart from the Net, some related data may also be available in local documents stored on your machine under various applications such as word, e-mail and the like.

Obviously, a search process that takes care of local data sources also — along with web and other Net sources — will enrich the search output quite a bit.

The new search tool, recently rolled out by the company, blinkx (http://www.blinkx. com/), is a product that attempts to address this requirement.

Once you start blinkx, which is available for free download, you will get its search interface, where you can enter the search string.

In response to your search, the tool scans its database and delivers the links with relevant data. Apart from displaying links from several Net sources, blinkx scans local directories specified by you and presents all the relevant digital files stored on your hard disk as well. If you want to search only your local storage, blinkx client software can be instructed to do so.

In addition to the properties mentioned above, the feature that makes this search tool a really innovative one is its ability to learn from the content of your current document and gather, display links that are related to the materials being viewed. That is, if you are reading a document on `peer-to-peer' technology, blinkx automatically produces several links related to this subject. The whole search exercise happens in the background, on the fly and that too without any intervention from your end.

To get the links displayed, you just have to move your cursor over that part of you application's title bar, where the small blinkx toolbar is displayed. As mentioned earlier, the results are displayed from various sources such as local documents, news web sites, web, video clips and web logs. By selecting a part of the text, you can restrict blinkx to provide results that are relevant only to the selected text.

This pro-active search feature of blinkx can be used in several situations. For instance, if you are composing an article on some theme and at some point if you look at the blinkx toolbar pasted on to your application, you will be able to see the links suggested by the tool. It is very much likely that some of them are relevant to the content of the article being composed.

Likewise, while you are on a web page, blinkx quietly reads the content and based on this feedback, generates links supposed to be relevant to the page's content.

Of course the links gathered by the tool may not always be relevant to the subject being discussed. When this author tried blinkx, many irrelevant links crept into the output along with a few useful ones. However, the concept is quite interesting and blinkx, which is in the beta stage, might improve over time.

Generally, when we get a search output from Google, we first visit the site listed on the top then move on to the second, third and so on. This navigation process is a little time consuming as you keep the browser idle while reading a web page. Actually you can speed up the process to some extent, if the browser can be made to automatically load the lower ranked pages while you read the ones at the top of the list. If you are a Firefox browser user, you can install the browser extension, LookAhead, to accomplish this. After installing the extension, LookAhead, whenever you want to conduct a Google search in the LookAhead mode, choose it by clicking on the Firefox's search bar. You can also select a word/phrase from a web page and invoke a `LookAhead' search process directly from the browser with this word as the search string by right-clicking on it.

For more details: http://lookahead. mozdev.org/

RSSCalendar

We know that XML-based RSS format files, though developed for publishing condensed information on the latest content posted on dynamic web sites, can also be used for many other purposes. This brilliant tool, which has changed the way many netizens gather information from the Net, is fast spreading its wings on to other applications. As mentioned in an earlier edition of this column, it is being deployed for diverse applications that include job search and learning object repositories. Recently this author came across another RSS-based service called RSS Calendar (http://www.rsscalendar.com/rss/) that enables you to create on-line calendar with RSS feeds. After signing-up with the service, select the `Calendar' option and create a web-based event calendar, where you can attach any number of events to specified dates. Once the on-line calendar is built, you can move over to the next step of RSS channel creation. At this point, the service will provide you with an RSS link, which can be used with a news aggregator by your colleagues/friends for getting the latest content available on your calendar. Anyone who subscribes to this calendar feed will automatically get updated with the latest information on your calendar.

On-line computing tips

While working with computers, you may often stumble upon new concepts and problems related to a variety of subjects that include software, hardware, networking and so on. Though all your questions may not get immediate answers, you can find many on-line help services that host lots of useful materials on various computing-related issues. The on-line computer support service, AnswersThatWork (http://www.answersthatwork.com/) is one such tool worth a visit.

Yet another on-line resource worth a look is the `Object Oriented Programming Web' (http://www.oopweb.com/index.html), a community web site built for programmers and students of computing. Here you will find lots of materials on a range of computer science/programming related subjects that include Operating Systems, Compilers, Java, XML, Perl and so on.

J. Murali

email the author at:
murali27@satyam.net.in

© Copyright 2000 - 2009 The Hindu