Badda Bing – not so Boom (well, yet anyway)

A BBC tech blog post in July indicated that there was a buzz about Bing, perhaps, but for Enlighten and our Theses Service, Bing is still a distant second to the search juggernaut that is Google – <caveat>according to Google Analytics</caveat>.

Bing – the “decision engine”

Bing (I am thinking Chandler not Crosby), Microsoft’s latest incarnation of its Internet search engine (or is that decision engine) was launched in early June this year. It’s focus is on 4 key sectors (shopping, local businesses, travel and health) [1] and MS has used it as an opportunity to pull together various other apps such as Multimaps into the Bing space.

The BBC blog entry notes that:

“84% of all searches in the UK in June were done with Google while Bing had just 3.3% of the market.”

For comparison, during June, Google Analytics recorded 8,139 visits to Enlighten and our first registered Bing hit was the 6th of June (and it was just one).

Bing as a referring site – the difference a couple of months make…

During June, Bing ranked very high in our referring sites category and we are not alone in this, Jenny Delasalle has blogged about this on the WRAP Blog .

Bing as a referring site in June 2009

Bing as a referring site in June 2009

In June, Google Analytics blog reported the advent of Bing and indicated that they were working on an update to manage Bing as a search engine rather than a referring site. The post provided a local fix for users who couldn’t wait for the GA update to filter through, we didn’t apply this and looking at data for the last couple of months this is now being done.

Bing as a referring site in July-August 2009

Bing as a referring site in July-August 2009

During July and August though Bing ‘referrals’ dropped from 2nd place during June as a referring site to 92nd during as Analytics caughtup and started properly treating access via Bing as search engine hits. Bing search engine traffic has grown to 4.48% of our overall searches, a distinct second to Google but well ahead of the rest of the “search engine” pack and the OAI search services which don’t really register at all.

Search Engine Access during July-August 09

Search Engine Access during July-August 09

As you can see Analytics shows that over 90% of our access comes from Google with Bing a distant second and Yahoo and even more distinct third.

Enlighten’s ranking in Bing

For “Enlighten” as a search term, we seem to have dropped to 3rd place on Google (from second during July/August, we have been overtaken by a dictionary definition of “Enlighten”). In Bing we are now back 11th place (where we were in July/August after a brief period in the top 10). This puts us on the 2nd page of results – is that the same as being invisible?

A more useful comparison is a search for “Language in Pictland” one of our most popular downloads. A comparison using bing-vs-google shows that both search services list the Enlighten version first, however Bing’s link is direct to the PDF.

Language in Pictland in bing-vs-google

Language in Pictland in bing-vs-google

Search Engines Arms Race

It’s still early days for this “new kid on the block” (if you don’t count its previous and varied incarnations) and while MS execs may be very excited about its overall performance, competition in the search engine arms race can only be a good thing – can’t it?

Innovation is certainly behind the recent Microsoft and Yahoo search deal which will see Yahoo Search become “powered by Bing” and enable them, they hope, to take on Google. The announcement of a new visual search service for Bing is an indicator of that innovation. We haven’t tried it yet though since you need to set Bing to be US and download MS’ Silverlight.

Google are not standing still though and have been working on “caffeine”, their new search engine “which will picks up news stories and puts fresher content higher up the search results” [2]

For the open access/repository community, does it matter how our content is found, as long as it is found?

The most surprising, perhaps disappointing thing about our Analytics reports is how little, if any traffic comes from the OAI search services like OAIster and Intute’s Institutional Repository Search, they just don’t seem to be on our users radar.

The challenge for repositories especially hybrid ones like Enlighten is to ensure that we can provide “discovery to delivery” rather than “discovery to digital dead end”. We need to ensure we can provide links to open access versions of our material and take advantage of tools like openURL resolvers to help our users.

[1] Microsoft’s New Search at Bing.com Helps People Make Better Decisions
[2] New Google ‘puts Bing in shade’

Advertisements

New journal fields added to Enlighten

We have added three new fields for the journal article document type for Enlighten. These are:

  • ISSN (Online) [Text]
  • Journal Abbreviation [Text]
  • Published Online [Date]
New Journal Fields in the record display

New Journal Fields in the record display

Example record: Multiple categories: the equivalence of a globular and a cubical approach

Autocompletion enabled

We have also refined journal autocompletion so that it will now fill
in five fields:

  • Journal title
  • Journal abbreviation
  • ISSN [Printed]
  • ISSN (Online)
  • Publisher

This is triggered by an entry in any one of the fields except ‘publisher’.

Journal autolookup example for title

Journal autolookup example for title

New searches

We have added ISSN, ISSN (Online) and journal abbreviation search options to our  Advanced Search screen.

New journal fields in Advanced Search

New journal fields in Advanced Search

A search for adv math or adv. math. in the abbreviation field will retrieve matching records.

Implementation

These new fields were added using the web Admin interface and then added into the workflow default.xml. New scripts were added to lookup to implement this in the new fields. We also removed the local stage which EPrints creates when new fields are created.

We also needed to add phrases into render.xml for the fields when they were added to the record display.

Enlighten’s 1,001 Multi-disciplinary Tweets

EnlightenPapers has now reached 1,001 tweets for on Twitter, that is, 1,001 new records since we added the Twitter code on the 9th of June this year. These include, among others, research in Celtic, English, History, Classics, Life Sciences, Law and the Physical Sciences.

The 1,001st Tweet

Our 1,001st tweet is a self-deposited full text paper from the National e-Science Centre (NeSC) at the University of Glasgow, “Applying formal methods to standard development: the open distributed processing experience” by Professor Richard Sinnott.

The 1,001st Tweet

The 1,001st Tweet

Twitter Growth and Visualisations

Twitter has provided an interesting gauge for our growth which we hadn’t anticipated when we started using it, for both the cumulative count of new additions and the rise (and fall) of followers. Our followers, currently some 65 [but it has been up to 80+] seem to be fairly volatile, many joining as papers match their interests and then leaving as perhaps they realise the volume of outputs and the broader range of material is not what they want as part of their own Twitter feed.

Twitter visualisation apps like Visible Tweets also provide us with new opportunities to showcase latest additions, and while such apps could be dismissed as party tricks they provide a glimpse of the range of re-use and visualisation options open to us. Visible Tweets also supports Twitter’s range of operators so that views can be refined by date, sender, hashtag and more. For EnlightenPapers we can use the limit from:EnlightenPapers to display only our own tweets rather than any replies which feature us.

Example of paper displayed in Visible Tweets

A paper displayed in Visible Tweets

Try EnlightenPapers in Visible Tweets – with rotation!

New Records, Bibliographic Services Staff and Training

Records, and freely available full text, have been added from a wide range of subjects over the summer months as we work with departments to add both their retrospective material and new publications.

Enlighten staff in our bibliographic services department have been very busy dealing with this increased influx of material, reviewing records, adding subject headings and checking copyright. They have also been involved in training sessions for departmental staff.

Since the beginning of 2009 we have run training sessions for staff in over 30 different departments about Enlighten, the University’s Publications Policy and the deposit process. These have been a mix of Powerpoint, hands-on work and coffee/tea; a formula which has proven to be a successful way to deliver the training and more importantly to start to build a “deposit community”.

Repository Growth – A Snapshot from ROAR

Using our entry in ROAR, we can track our growth since 2004 when the Enlighten [formerly the Glasgow ePrints Service] was launched as part of the JISC funded DAEDALUS project. We have had steady growth but this is now really starting to accelerate as the training and publications policy start to make themselves felt. We have already added some 2000+ records since the beginning of 2009 and more than half of those have been added in the last 3 months.

Growth of Enlighten since February 2004 from ROAR

Growth of Enlighten since February 2004 from ROARMAP