Material selection is one of the most important tasks for a library. LSC has the expertise to provide a superior selection service that will result in a high quality collection.
LSC has the expertise to assist the library in creating profiles that will drive circulation. Since LSC is a “not-for-profit”, our selection advice is not driven by publisher considerations: we truly are an extension of the selection expertise that the library has.
Budget management is a crucial part of any ARP process. LSC’s integrated system links budget lines, orders, price changes, and shipments to provide an accurate ordering and shipping process. When there are changes – such as the Canadian dollar moving – LSC will react in a timely manner not at the last minute.
LSC uses a combination of deep data mining and staff selection expertise to provide ARP selection.
Data mining is the process of looking deeply at all of the available data in order to drive objective recommendations. It differs from artificial intelligence based solutions because it uses the data to drive the types of further iterations: artificial intelligence seeks to create rules that can be automated at every iteration.
Superior data: any approach to apply evidence to collection rules requires data to work. The key source of this data is the circulation data from the library’s ILS. The core value that LSC adds to this data is that we link it with the actual cost of the material. Using tools like CollectionHQ, LSC is able to apply our comprehensive data based on the Canadian market. Where the library is able to provide ISBN level circulation data, LSC is able to become even more precise in our analysis.
In either event, LSC’s analysis is based on the profile of the library.
Within the framework created by the data, LSC draws on our staff selection expertise. We have one of the largest and most experienced selection teams in Canada. They do more selection for more libraries than any other selection team in Canada. This provides them with a depth of experience not found with all suppliers.
There has been some development work done to promote artificial intelligence as a cost effective tool for selection. This approach correctly identifies the need to use data to drive key areas of collection resource allocation. However, it suffers from some significant flaws:
Artificial intelligence is a set of complex computer rules that allow for iterative “learning” of optimal solutions based on experience. For example, when to lower prices during the Christmas season: a system can test solutions and then “learn” when to make the necessary changes. Continuous feedback allows for adjustments as the environment changes.
But this is fundamentally a backward looking approach that is ill suited to selection of individual titles. A library can never pick all of the available titles. System driven selection will consistently recommend titles based on past overall performance in the data sets that are used.
If there is no or limited use of data in the development of selection criteria, then the selection will veer towards the personal biases of the selection staff. The library will have a great collection in those areas that are favourites of the selection staff or those in control of the collection profile. But the library collection will not evolve to meet the needs of the patrons as they change.
Many public libraries have an adult non-fiction collection that is over sized based on its actual use. Non fiction works are seen as important and part of what a library “should” have. This leads to an over allocation of materials expenditure in this area.