Robustness of IR Models to Collection Growth
2026-08-24 • Information Retrieval
Information RetrievalComputation and Language
AI summaryⓘ
The authors studied how search systems find relevant documents when new, unrelated documents are added to a collection. They wanted to see if adding non-relevant documents makes the system worse at finding what you want. They grouped systems into two kinds: those that consider other documents in the collection when ranking (MDD) and those that do not (MDA). Their experiments showed that both types get a bit worse when irrelevant documents are added, but MDA models generally handle this situation better in retrieving documents. For reranking results, both types performed about the same.
Information RetrievalBM25IDFMulti-Document-AgnosticMulti-Document-DependentDocument RankingRerankingRobustnessDynamic CollectionsPerformance Degradation
Authors
Emmanouil Georgios Lionis, Debasis Ganguly, Sean MacAvaney
Abstract
Information Retrieval (IR) systems seek to identify relevant documents within a collection. In practical applications, collections are dynamic, with documents frequently added. We argue that ideally, a retriever's effectiveness should not decrease when non-relevant documents are added to a collection. This study formalises this concept and empirically evaluates it by merging two collections with negligible topic overlap. We hypothesise that the way an IR model conditions its ranking on other documents in a collection (e.g., the IDF component in BM25 or contextual documents in listwise rerankers) plays an important role in its robustness to the addition of non-relevant documents. We broadly classify models as those that do not depend on other documents (Multi-Document-Agnostic, MDA) and those that do (Multi-Document-Dependent, MDD). Our results show that neither MDD nor MDA models are fully robust to the addition of non-relevant documents, as all models exhibit some performance degradation. Interestingly, among the models we test, MDA is more effective than MDD for retrieval, whereas MDD and MDA rerankers are equally effective.