A decision support system for optimizing book pickup and re-shelving in libraries

As the number of students in universities increases, the number of books borrowed and used in libraries also increases, leading to more books returned and re-shelved. Library personnel need to walk greater distances within the library and their work efficiency decreases. This thesis proposes a Decision Support System (DSS) to improve the re-shelving process. Different book pickup and re-shelving scenarios for libraries are studied as a network optimization problem in the DSS. Tables and shelves in the library form nodes and walking corridors form edges. This problem is essentially a sequential ordering problem, with the objective of minimizing the total distance traveled and relationships of priority between nodes to be visited. Scenarios are solved by using different heuristic and meta-heuristic methods. These include particle swarm optimization meta-heuristic in addition to the nearest neighbor and the sweep algorithm heuristics which are commonly used for similar problems. Additionally, a specific heuristic is proposed for the problem examined in this thesis. The performance of different scenarios and solution methods is compared using randomly generated problem instances. It is shown that scenarios involving book pickup from just one or two locations within the library are significantly more effective than others. The book pickup and re-shelving optimization problem considered in this thesis has not been studied in the literature before. In addition, the DSS utilized to solve the problem is the first DSS to provide features including a library database, network structure, book pickup and re-shelving scenarios, solution methodologies, and solution comparison. For this reason, it is believed that the thesis will make important contributions to both the literature and efficiency issues in libraries.