Wednesday, November 14, 2007

Improving Web Browsing for Legal Research

Alert law librarians may have noticed that Westlaw has introduced a beta version of a proprietary web browser, WebPlus. You can see (and hear) a free demo and tutorial without logging onto Westlaw here. Released to public use in mid-July, the product is free to users right now. I understand that Westlaw is trying to work out a model where they can integrate the results of a WebPlus search with a standard Westlaw search, similar to how they produce Results Plus materials in a side-bar now. In the alternative, they want to include Results Plus materials in a side-bar to a WebPlus search. Their problem right now is to work out how to charge for searches on the Westlaw databases and not charge for web browsing results, while still integrating the two search results. If they can figure this out, I think WebPlus will be a useful adjunct to Westlaw searches.

WebPlus seems to be designed to neutralize "search engine optimizers" (SEOs; see this article on Wikipedia, with links). Many librarians are aware of the arms' race between web designers and search engines. The reason Google keeps much of its ranking algorithm secret, for instance, is to reduce the gaming web designers can do to promote clients' webpages in the Google ranking. But then, they sort of undercut those efforts by offering "sponsored links" to Google searches. If you search the term "search engine optimizer," you will find loads of designers and consultants to claim to improve your web pages' ranking, and therefore bring more eyeballs to your site. WebPlus is trying to reduce the ranking of purely commercial websites that have no useful content, partly by ignoring the "sponsored links" results. Likewise, law firm web pages that are just offering services are ranked lower by WebPlus, promoting the pages that carry real information.

WebPlus is not so much a search engine as a ranking system, designed to make results more focused and relevant for legal-oriented searchers. The ranking system is the key improvement. I understand that they are searching as widely on the Web as Google, say, but the results returned to the user promote the results of the web pages most likely to be of interest to legal users. Westlaw editors, each assigned a subject area are now tagging web pages and blogs with KeySearch keywords. They are selecting websites that are key and information-rich in their subject specialty. So, there is a major aspect of this new service that is organizing and indexing the Web. They are also crawling the web with Microsoft Live, and then ranking the results in real time.

The WebPlus system is also a learning machine which follows which results a user selects in order to improve its future searches' relevancy. It is set to weed out (rank lower) purely commercial websites, and to focus on information-rich sites instead. A good way to test WebPlus results is to compare the first page of results with the first page of Google results. For instance, you can try out a search one day and try it again a few days later, and see different results -- partly from the selections of other WebPlus users. Relying on a cadre of like-minded users over time, their choices should be improving the relevancy of the ranking system. Try a search of "aliens" or "RICO" on both WebPlus and Google, for instance, and see if you get fewer spaceship sightings or other false drops in WebPlus.

WebPlus allows a user to set "context settings" to focus on legal (the default setting), people, companies, government, news or the general Web. Westlaw told me that they were trying to respond to high use of the Web by associates, often using very short, unsophisticated queries that retrieved bazoodles of results with very low precision. They also hope that this additional link to the Web will help fill the "law and X" gap that is looming since the Dialog databases left Westlaw. Eventually, I think, Thomson plans to introduce a version of WebPlus for science, business and health as well.

Of course, there are also excellent guides on improving Google searches such as this excellent article by Tracy Rich at National Law Journal's

1 comment:

Victoria Szymczak said...

I am wondering if the West web search engine might be part of the reason why they, and LX, are not playing with Serials Solution and other services providing federated searching and the like. I understand there is some concern because they are not updating w/ ss as quickly as we would all like. I won't be surprised if we see a LX competitor to the West product. I think the West product is pretty good, but I think the Cornell legal research engine covers legal free web sites pretty good already.