Microsoft have launched their new look for Live.com. It has the now obligatory minimalist look, which was already evident in the previous interface, but has at long last added a link to the Advanced Search option on the home page. Also new to this version is the option to receive an RSS alert for news searches. “Hurrah!” I shouted, “At long last they are listening to users and in danger of threatening Google’s crown”.

Unfortunately, my joy was short lived.

1. The advanced search screen is still pathetic compared to Google’s, Yahoo’s and Exalead’s, and there is no filetype search option. You have to use the ‘filetype:’ command in the default search box

2. A major issue I have had with Live is that it offers different search options and results displays depending on which ‘country version’ you are using (see the slide below from one of my recent presentations comparing the UK and US versions and number 3 on the News search).

Live.com: UK vs US versions - March 2008

In the previous version of Live.com you could force it to switch from, for example, the UK to the US version by going into the Language option and choosing English (US) instead of English (UK). Now, there is no differentiation between US and UK. I thought I might be able to solve this problem by going into Options and changing the location at which Live thinks I am based. It assumes London but even when I tell it that I am in New York, United States it still insists that I am in the UK! A minor issue you might think but if your ISP gives Live an IP address in Frankfurt, Australia or wherever and Live is telling you that it is going to give you customised results according to your location - well, what is the point? Google and Yahoo give you the option to switch between different country versions whenever you want.

3. At long last they have implemented RSS feeds for news search alerts, but then I realised that I was looking at the boring old news results for the UK and not the super-duper display that the US now sees (see my earlier posting on this issue). And there is no way that I can find, other than going through an anonymous proxy server based in the US, to gain access to the US version.

The verdict? I have to partially agree with Phil Bradley’s comment “If I get a delegate on a course asking me why Live Search should be their preferred search engine I simply couldn’t give them a good reason.” They have so much going for them and then they totally mess it up. Their database is the most up to date for many of the sites that I search on; the coverage seems to be better; they have a worthy competitor to Google Scholar in Academic Live; Maps, Books and Live Earth are pretty good too. So why do they keep shooting themselves in the foot with c**p interfaces?

The presentation that I gave at the Association of UK Media Librarians on 19th April 2009 is now available. You can find it on my own web site as a PowerPoint (8 MB), on Slideshare and authorSTREAM. As usual, it probably won’t make much sense if you were not at the presentation but at least it gives you an idea of the tools I was highlighting.

PharmaLive Search (http://www.pharmalivesearch.com/) is maintained by publisher PharmaLive and powered by Convera. It searches PharmaLive’s collection of publications, selected industry and therapeutic web sites, and relevant blogs. You can search Publisher Recommended Sites, PharmaLive.com, or the Web. Although the web search option is not limited to specially selected sites, they are ranked for relevance with respect to the subject and have been filtered for spam.

‘Publisher recommended sites’ covers over 25 000 000 documents from more than 2600 selected domains. ‘Search PharmaLive’ provides results from the PharmaLive itself and its print publications. Much of the information is free but articles from PharmaLive’s own publications are priced.

Variations in spellings, for example hypoglycaemia and hypoglycemia, are automatically included in your search as are synonyms. My search on mushroom poisoning picked up terms such as toxicity, and ‘adverse reactions’ picked up ‘adverse effects’. As one would expect from a vertical search engine, the synonym identification is vastly superior to Google’s efforts. If you make a complete mess of the spelling it automatically prompts you with a “did you mean..” option.

At the top of your results page there are links to broader, narrower and related terms, and to the left of the screen are displayed most popular searches and related concepts. You can focus your search further by using the category tabs. For example, if you have chosen to search publisher recommended sites there are tabs for All, Associations, Media, Government (as far as I could see this covers relevant bodies world-wide), Companies, R&D, and Blogs. When searching PharmaLive.com you are offered options for Med Ad News, R&D Directions, and Pharma Live News. Those articles that are only available on subscription are marked with a blue padlock.

If you regularly search for pharmaceutical and health care information , add this one to your list of tools and also consider downloading the PharmLive Search toolbar.

TripleMe.com is another tool that allows you to run a search across multiple search engines and display the results in separate columns on one page. It covers Live, Google and Yahoo, and there is a fourth column for Google ads. Phil Bradley mentioned on his blog that before he could his first results he had to email a friend and tell them about it. That particular ‘feature’ seems to have gone.

Phil Bradley was not very complimentary in his blog about Tinfinger, a people search tool. His search on Gordon Brown, the current UK prime minister, came up with an entry that had clearly been computer generated and which was full of inaccuracies. Tinfinger’s Paul Montgomery confirmed in a comment to Phil’s posting that a lot of the content is pre-generated, but that the intention is for people to use that raw data to write full profiles.

It seems that Phil’s posting encouraged someone to update and correct Gordon Brown’s profile as the information is now correct - or was when I looked at it. Tony Blair’s is a different matter, though. There is no profile but there is a picture and several “tags”:

“Blair has the name of The Rt Hon. Tony Blair. Blair has the type of Alumni of the Inns of Court School of Law. Blair is on the Sedgefield (UK Parliament constituency) list. Blair is listed with the external id of 0086363. Blair came before “(constituency created)“. Blair is an devotee of the Anglican religion. Blair started in 1983. Blair was preceded by John Major. Blair’s spouse is Cherie Blair. Blair has the order of Prime Minister of the United Kingdom. Blair’s term started on 1997052. Blair represents the constituency of Sedgefield (UK Parliament constituency). Blair’s deputy was John Prescott. Blair’s job is politician.”

I would love to know from where that information was collected! It is early days for Tinfinger and hopefully people will get on board and start editing and creating more accurate profiles. At present it is an example of how wrong information can be when it is automatically generated, and the importance of being aware of how content is compiled.

Yet another gem discovered via Phil Bradley’s blog. FuzzFind Web Search is a meta search tool that combines Google, Yahoo and Live searches with results from social bookmarking site Del.icio.us. The results are amalgamated and deduplicated into a single list. So what? Meta search tools are 10 a penny. But I really like this one because of the way it shows me where each result has been found in each search tool by placing icons and numbers to the left of each entry in the results list. I don’t know why that particular and very simple feature impresses me but it does. Search, and the presentation of the results, is a very personal thing and what works for one person will not necessarily work for someone else. This one really works for me.

Zuula is another search engine along the lines of Intelways and Trovando. You type in your search once and then run it through several search engines one by one. The search engines are grouped into Web, Images, Video, News, Blog and Jobs. It is not as wide ranging or as comprehensive as Intelways, for example it does not have a group for searching by file formats, but it does offer an Advanced Search screen that includes a domain search. You can also keep a list of your previous searches.

Chipwrapper is a Custom Google Search Engine that searches across the UK’s major national newspapers: The Daily Express, Daily Mail, Daily Mirror, Financial Times, The Guardian, The Independent, The Sun, The People, News of the World, The Scotsman, Daily Star, The Telegraph and The Times. It also searches the BBC News web site, ITN and Sky. It carries out a Google web search of these sources, not a Google News search, so although you cannot sort the results by date you do pull up older, archival material that is not included in the standard 30 days of Google News. As many of the newspapers give the date as part of the text of the stories you can often limit your search to a year and sometimes a specific month by simply including the month and year in your search. Also available are RSS feeds for the top headlines, general sport, rugby and football headlines, and a Chipwrapper browser search plugin is available for IE7 and Firefox 2.

I ran one of my standard test searches on “Richard Budge“, the UK coal magnate, and included 2007 in the search strategy to limit the stories to the current year. It worked impressively well but the FT was conspicuous by its absence, and I knew that there had been at least two articles about him in the FT this year. I went direct to the FT site, ran a similar search and found three articles. I then did a Google Web search using my Chipwrapper strategy but added site:ft.com. Again three results. Back to Chipwrapper with some different searches, and it did pick up FT articles. Obviously there is something about my Richard Budge search that it does not like.

Comparison with Google News Archive

Of course Google News is probably the service that most people will use as the benchmark and this is where it becomes really interesting. Back to dear old Richard and a search in the Google News Archive. I went into the Advanced Search screen, entered the phrase Richard Budge, typed in 2007 for the year and selected Show Timeline. Unlike the Advanced Search in current Google News you cannot specify a country of origin for the source so I had to resign myself to the possibility of wading through a substantial number of articles. Google Archive News does, though, give you an option to home in on a specific month via the Timeline (see below).

Google Archive News Results with Timeline

It came up with 94 results, about twenty of which claimed to be from the Financial Times and dated Jan 1 2007. I clicked on a few of the links and they took me to the “Access My Library” site where I was repeatedly told that the articles had been deleted. I gave up after eight or nine, but I think we can assume that the FT has decided not to play ball. Three links with $$ signs next to them took me to Press Display but I was told that the items had been “removed from the back issues access”. A minority of the links took me direct to the news source, for example The Telegraph, BBC, Doncaster Today. The articles from the Guardian, Independent and Times that had been picked up by Chipwrapper were nowhere to be seen. I can only think that like the FT they have declined Google’s offer to be part of the Archive.

Conclusion

I have to confess that this is the first time I have analysed the results from Google News Archive in any depth. I was not surprised to find the FT absent but amazed that so many of the other UK daily papers were not there. Even worse, Google still has in its index links to stories that were carried by third party services, such as Access My Library and Press Display, but which have now been removed. The Timeline is still a good way of looking at major stories relating to a company or person but be aware that some of the key resources are not included.

Despite the glitch with the FT in my test search, and that there is no date sort option, Chipwrapper is a great tool for searching new and archival stories appearing in the leading UK papers. I recommend that you give it a go next time you need to research a UK story.

Note on the name Chipwrapper for non UK readers

A favourite take-away food in the UK is deep fried, battered fish with chips (mushy peas are optional but an essential component as far as I am concerned). Today, EU and Health and Safety regulations dictate that these have to be placed in grease proof paper (in practice not grease proof at all) and then wrapped in large plain sheets of off white paper. I recall that in my dim and distant youth newspaper was regarded as the superior wrapping material. Some connoisseurs claimed that the ink, which dissolved in the presence of the salt and vinegar, gave extra flavour.

As Greg Notess and others have already noted, Live.com’s link and linkdomain commands are in a mess again. After they had been disabled in their original form, they reappeared as +link: and +linkdomain: commands. I noticed last week, though, that the +linkdomain was generating some very strange results. Much as I would like to believe that I am very popular I do not honestly believe that over 500,000 people/pages/sites link to my web site! Yahoo’s result of 2895 seems more realistic. Now Live’s commands have gone AWOL again.

Using Live’s Advanced Search screen I can still use the links option, type in the URL of the page and the syntax that it comes back with is link:http://www.rba.co.uk/. But that only gives me 7 results and two of those are internal links on my own site. So I guess it is back to using Yahoo for identifying incoming links.

Live’s ‘linkfromdomain:’ is still working.

AlltheWeb Livesearch seems to have disappeared. Clicking on the link to Livesearch on the AlltheWeb home page or trying to go direct to http://livesearch.alltheweb.com/ causes both my Firefox and IE 7 browsers to grind to a halt and then redirects them to Yahoo. Does anyone know if Yahoo have decided to abandon it, or is it a temporary glitch?