As both technology reporter and consumer I rely heavily on web search tools to find and then track through the thickets of information accessible to anyone with a PC and internet connection.

My browser home pages point to Google, or more precisely iGoogle, the personalised version, which I have set up to display an array of news, blogs and other data. But that does not mean Google – and any of the other general purpose search engines – are always the best tools for finding information, especially if you are interested in tracking trends, opinions and relationships. For this you need to tap into next-generation “semantic” search tools, such as Hakia (www.hakia. com) that try to understand concept of a search query using techniques such as sentence analysis. Most other leading search engines rely mainly on analysing keywords.

So I got excited at the Demo technology start-up show in Desert Springs, California, last week when I came across several interesting new search tools and other neat new technology (www.demo.com; www.ft.com/techblog ).

Two search tools that caught my eye were Silobreaker and Toms (Top of Mind Service) from Jodange (www.silobreaker.com; www. jodange.com). They both use sophisticated algorithms to deliver the sort of valuable insights that professional researchers or the simply curious crave.

UK-based Silobreaker is a news and current affairs search engine that tries to look at the data it finds in the way a person would, teasing out relationships and putting companies, people, topics and places into context. Silobreaker’s creators say it pulls content on global issues, science, technology and business from about 10,000 news, blog, research and multimedia sources.

It can graphically display “media attention trends” and compare people, companies or products to see who or what attracts most media attention. For example, you can track and compare the media “buzz” around the US Democrat candidates over the past week, month or year, track the attention paid to rival internet services such as My­Space, FaceBook and You­Tube, or find the key players in the Middle East peace process.

It can also chart article volumes by media type, identify geographic news hotspots and, impressively, map relationships between topics in the form of an interactive network. Users can navigate spiders-web graphics by dragging and dropping the elements in it. I found the graphic often suggested relationships I had not immediately thought of.

While Silobreaker tracks down salient quotes as part of its analysis, Jodange’s Toms is a “sentiment analysis” engine that focuses entirely on opinions and appeals to users in the securities industry and financial services sector. For example, it answers questions such as what chief executives think about subprime lending; which companies mentioned research and development in their earnings reports; or what Apple’s Steve Jobs says about innovation.

The technology behind Toms was developed by Cornell University computer scientists and uses linguistic analysis to extract opinions from documents and identify the opinion holder and topic. Toms analyses hundreds of thousands of documents tracking Fortune 1000 companies and S&P 500, including financial reports, transcripts of quarterly calls, executive presentations, analyst reports and financial blogs. It aims to identify who is worth listening to and how best to correlate opinions to outcomes.

For example, users can track what a particular stock analyst has said about a company and what impact that had on the share price, or track dozens of experts on oil prices or interest rates and see when their opinions change.

Delver (www.delver.com) takes a different and more consumer-focused approach to search. Its starting point is that traditional search engines’ lack of information about you means the results they deliver are pretty much the same whoever types in the keywords.

Delver aims to deliver more relevant search results by tapping into a user’s friends and online social networks. The service analyses your friends’ tags and information found on social networking profiles, blogs, bookmarks, photo and video sharing sites and tailors search results accordingly.

For example, Delver searches your public profile on sites such as Facebook, Flickr, and YouTube to discover who your friends are. Then when you enter a search term, say “Indian restaurants”, the results should include a list of restaurants your friends have recommended or have at least talked about. Delver has just launched as an invite-only “beta” trial.

The other search-related start-up that drew my attention at Demo was Iterasi (www.iterasi.com), which addresses a problem familiar to almost anyone who has bookmarked a web page – typically bookmarks save a link to a page’s location rather than the content so when the page changes the information is lost. Iterasi is a browser-based add-on tool that captures and saves the content of a page as it appears at that moment so it can be retrieved with one click.

Once the toolbar is installed (it currently works with Internet Explorer 7 only, although Firefox and Safari versions are promised) you click a button and a copy of the entire page is transferred to a secure personal account.

Users can access their Iterasi account from any web browser. Iterasi power users can automatically capture pages on a regular basis so they can track and analyse changes over time. Like Delver, Iterasi is currently on invitation-only trial, but the developers hope to launch this simple but smart free service by mid-year.

paul.taylor@ft.com

Paul Taylor tackles your high-tech problems and queries at www.ft.com/gadgetguru