LawSource, Quicklaw and CanLII all permit users to find documents by name or citation, as well as to conduct keyword searches in the full text.
LawSource, Quicklaw and CanLII assume that most users will want to search the entire case collection. However, users can easily filter or refine the search results, by jurisdiction, court, date, citation frequency, and various other filters.
Federated searching and integration
In the new WestlawNext Canada platform, the default is a federated search of all collections (except finding tools), using the plain language search engine. You can simply type in a question, or a string of words, without connectors.
The federated search includes commentary as well as primary sources. It is more likely that users will view commentary earlier in the research process because of this. In addition, there is effective integration of primary sources and KeyCite with some of the finding tools and commentary.
A search can be done of a discrete collection, by selecting the collection first from the home page and then running a search. This is of real benefit only if the user wants to search the collection using a template, or if the collection being searched is one of the finding tools listed at the bottom of the home page.
The federated search also includes results from material that is outside of subscription. Those items are clearly marked, and can be viewed for an additional charge.
Plain language searching
The plain language search engine looks for various word forms of your search terms, carries out phrase recognition for legal and general use phrases, and finds equivalencies for some commonly used acronyms. It relies primarily on frequency and proximity algorithms to rank documents. It also uses citation relationship data from KeyCite Canada to help assess relevancy based on relationships among cases.
The plain language search engine uses “best match” rather than finding documents that contain all of your search terms. This can help you find cases that are highly relevant, even though a term you included in your query does not appear in the case. However, sometimes you want to ensure that particular terms appear in the retrieved documents. Here are a few techniques to fine-tune plain language searches:
- use quotations marks to force a phrase search
- to force a word to be included, put a + directly in front of the word
- to force a word to be excluded, put a – directly in front of the word
- add synonyms to your query
- by simply adding them to the query, and letting the search engine sort it out, or
- directly after the relevant search term, add any synonyms within parentheses, with multiple synonyms separated by commas
Those who prefer Boolean searching can trigger that type of search by including connectors in the query:
- A query containing a proximity connector or a truncation command will automatically be processed as a Boolean search.
- If you include the words AND or OR in your query, or quotation marks, you will be asked if you want to run the search using Boolean syntax.
- Another option for running a Boolean query with the assistance of a template is to click on ADVANCED (by the search box). It is best to select a particular collection first, such as Cases or Legislation, so that a template customized for that collection can be used. A tab will appear on the search box to notify you that your search is restricted to a collection.
- If you switch to Boolean, remember that some of the advantages of plain language searching will be lost, such as automatic word-stemming and augmented relevancy ranking. In a Boolean search, any spaces between words will be processed as OR rather than AND.
- “advanced:” will appear in the search box in front of any query that is run as a Boolean search. This is a useful reminder of which search method is processing your query.
The plain language search engine works very well. However, it is worth understanding the Boolean syntax used on the WestlawNext Canada platform, because any refinements that you do using the “Search within results” feature will be processed as Boolean searches.
Search results, refining and sorting
The result list will indicate which items are outside of subscription. The result list is broken down by collection type. This enables different types of filters for each collection.
Because the plain language search uses a “best match” system, it can retrieve hundreds of documents and many of them will not contain all of your search terms. As a result, relevance can drop off quickly. This causes problems when you re-sort your results, for example by date. The most recent documents may not be very relevant.
It is therefore advisable — particularly for plain language searches — to use some of the filters to restrict your results, either instead of re-sorting or before re-sorting.
- Select the document type you want to review from the list on the left.
- Review the first several documents until relevancy drops off.
- Then use “Search within results” to narrow results using Boolean terms, before you re-sort. This should dramatically reduce the number of documents, and will show only the sub-set that satisfies the Boolean query.
- Alternatively, instead of re-sorting, use the filters to restrict the results by court level, jurisdiction, citation frequency, or date. The advantage of this method is that you won’t lose the original relevancy ranking.
A detailed record of each search session can be sent to you automatically by email after sign-off. This is useful for keeping track of queries and documents reviewed, and should be filed with your research notes.
The WestlawNext Canada platform saves your search history for up to one year. The history lists searches conducted, documents reviewed, and KeyCite results generated.
You can search within all of the documents you have previously viewed within the past year, and filter those documents by client ID or by search date. KeyCite flags on documents within the search history will automatically update to reflect new history or judicial consideration.
Annotating, saving and sharing documents
When viewing a document, the user can annotate the document by adding a note, and by highlighting passages. In order to keep the annotation, the document must be saved in a folder or downloaded with the annotation.
Users can create folders on WestlawNext Canada and save viewed documents, including any annotations, within those folders. As with the user’s search history, KeyCite flags on these documents will automatically update. The contents of a folder can be downloaded by the user at any time.
A folder can be shared by the user with other WestlawNext Canada password-holders in the same law firm. However, the annotations will not be shared; only the documents saved to the folder.
Some controversy has arisen because the annotations are stored on servers in the United States, and therefore subject to the Patriot Act. Consideration should be given to this when deciding whether to use the annotation feature, and the type of information to include. For further information, see the blog post on the Canadian Association of Law Libraries website entitled WestlawNext User Annotations Being Stored on US Servers.
Quicklaw uses Boolean searching for creating keyword search queries, in customized templates developed for each type of collection.
Quicklaw also offers a natural language option on the Home tab. This natural language searching option is less sophisticated than the plain language feature on WestlawNext Canada. However, if Boolean searches are not turning up relevant results, you may want to try a natural language search. You can force phrases (using quotation marks) and specify mandatory terms when you create the natural language search.
Quicklaw includes an auto-complete feature on the home page for name or citation searches, and for words and phrases searches.
An approach to searching that is based more on federated searching and natural language is available to QL users who subscribe to the Quicklaw for Microsoft Office add-on.
Search results, refining and sorting
Search results on Quicklaw can be re-sorted and filtered using various parameters. One of these parameters is a topical classification scheme. It can be used prior to running a search, to narrow the material being searched. However, a more effective use is to run the search, and then use the topical filter when you review the results.
A recent QL enhancement allows users to turn on an option that eliminates duplicates from case law results.
The History & Alerts tab in Quicklaw lists documents viewed over the previous 48 hours (up to 100), and searches run over the previous 7 days (up to 100). Links are included to facilitate viewing a document, running a search again, or saving a search to be run periodically that will alert you to new documents meeting the search criteria.
The CanLII search box can accommodate queries for document names, citations, or full text searches. If connectors are used, the search engine will recognize them. However, a string of words without any connectors can also be searched effectively. The CanLII search engine incorporates some natural language features such as word stemming and phrase recognition.
CanLII search includes an auto-complete feature that is triggered when you start typing the name or citation of a case or legislation. Another recent CanLII change is the enhanced ability to filter search results.
Currency and accuracy
CanLII is very fast to process new case law, because of its reliance on technology rather than editorial oversight to carry out various tasks, such as generation of keywords and links. As a result, new cases are often published on CanLII before they are available on either LawSource or Quicklaw. Using RSS alerts on CanLII may also result in faster alerts.
However, this reliance on automation can occasionally result in missing links, which means that some note-up results will be incomplete. Also, value-added features such as treatment codes are not available on CanLII.
See the comparison chart of search syntax for WestlawNext Canada, Quicklaw and CanLII.