Making sense of usage statistics

One of the buzz words floating around our industry over the past year or so has been SUSHI. This may evoke delicious images of nicely-presented plates of rice, seaweed and raw fish but it also has a more technical meaning.

Although the Standardized Usage Statistics Harvesting Institute (SUSHI) is less tempting for the taste buds than the Japanese food that shares its name, this National Information Standards Organization (NISO) initiative has a very important job in this industry. It aims to solve the challenge of how librarians can track usage of online content.

Retrieving usage statistics for a single journal is relatively easy but, for libraries, the process of having to visit publisher website after publisher website can prove to be extremely time consuming - particularly when you consider the volume of journals involved. And, of course, merely retrieving the statistics is just part of the process. The librarian then needs to evaluate those statistics and this can lead to more problems.

In some cases, the statistics retrieved from a publisher will be presented in the standard COUNTER format but, occasionally, the data is coded in the publisher’s own internal format. Sometimes the data is only available on a web page and cannot be downloaded, only printed. Having statistics in a variety of formats makes the task of evaluating those statistics as a whole incredibly complicated.

What SUSHI does

This is where SUSHI comes in. It aims to provide a protocol that will enable libraries to retrieve and analyse their usage statistics automatically. The SUSHI Draft Standard for Trial Use was released on 20 September 2006 and a PDF version of this draft can be downloaded from the NISO SUSHI Committee website. By default, this protocol forces the publisher to provide usage data in a standard format (COUNTER XML). The SUSHI protocol is then a SOAP (Simple Object Access Portal) request/response web services ‘wrapper’ for the XML version of COUNTER reports.

In this protocol, a transaction begins with the library’s client service. This can be either an internally-developed application or one running as part of a usage data consolidation service or integrated library system (ILS)/ electronic resource management (ERM) system. The client service identifies both itself and the customer whose statistics are being requested to the SUSHI server service that is running at a data provider. It also specifies what report is required. In response, the server provides a report in XML format, along with the requestor and customer information - or an appropriate error message.

SUSHI’s developers aim to reach a level where the client system is programmed to automatically retrieve reports on a monthly basis, for all COUNTER-compliant vendors with which the library does business. Following the memorandum of understanding that NISO and COUNTER signed earlier this year, the NISO SUSHI working group is now responsible for maintaining the COUNTER schema. This schema describes the internal structure of the XML files being delivered. It is included to enable the user to anticipate and build their applications with the knowledge of the structure of the incoming file.

Furthermore, although the SUSHI protocol was developed initially to handle COUNTER-compliant reports, it will evolve over time into a delivery mechanism for non-COUNTER reports and was designed in such a way that this can be done easily. Trading partners need only to agree on the format and to register their report name with NISO.

The benefits for libraries

The primary benefit of SUSHI is that it automates a tedious and repetitive process. SUSHI will be a tremendous time-saver for librarians. The fact that they will no longer have to spend all that time retrieving usage statistics on a publisher-by-publisher basis or worry about a number of differing formats will be ideal for them. For publishers, SUSHI is a tool that will allow them to standardise not only their usage statistics into a common format, but also the delivery of that information.

As far as the widespread adoption of SUSHI goes, we do not foresee any major obstacles that would hinder this because the protocol makes use of standard web services and the widely-adopted COUNTER reports. We have already seen a rapid take-up of ERM products in libraries. SUSHI is central to one of the promises of ERM in allowing libraries to more easily manage their electronic resources. Librarians that have invested in ERM systems that support SUSHI will not need to do anything extra to comply.

With ERM vendors already adding support for the standard, it is likely that libraries will be pushing for publishers to implement it. Publishers and librarians who are developing local solutions can find the SUSHI schema, as well as toolkits for the Java and NET environments, on the NISO SUSHI Committee website. And, where publishers have not had a chance to implement it yet, libraries can look to aggregators who are SUSHI-compliant for the same information.

Swets is one of the founding members of the SUSHI initiative. As an intermediary, Swets is constantly searching for ways in which we can help our customers manage their information resources in a simple and streamlined manner. Driving the efficiency of our working processes is key to providing that and as such we have added support for SUSHI to our SwetsWise online content service and encourage our customers to make use of this wonderful standard. In the end, SUSHI should help to make the delivery, handling and evaluation of usage statistics a clean, simple and transparent process for all involved. Just don’t try to eat it!

William Hoffman is process analyst of Swets Information Services and a member of the SUSHI committee