LLM-powered search agents are increasingly being used for multi-step information seeking tasks, yet the IR community lacks empirical understanding of how agentic search sessions unfold and how retrieved evidence is reflected in later queries. This paper presents a large-scale log analysis of agentic search based on 14.44M search requests (3.97M sessions) collected from DeepResearchGym, i.e., an open-source search API accessed by external agentic clients. We sessionize the logs, assign session-level intents and step-wise query-reformulation labels using LLM-based annotation, and propose Context-driven Term Adoption Rate (CTAR) to quantify whether newly introduced query terms are lexically traceable to previously retrieved evidence. Our analyses reveal distinctive behavioral patterns. First, over 90\% of multi-turn sessions contain at most ten steps, and 89\% of inter-step intervals fall under one minute. Second, behavior varies by intent. Fact-seeking sessions exhibit high repetition that increases over time, while sessions requiring reasoning sustain broader exploration. Third, query reformulations are often traceable to retrieved evidence across steps. On average, 54\% of newly introduced query terms appear in the accumulated evidence context, with additional traceability to earlier steps beyond the most recent retrieval. These findings provide candidate signals for repetition-aware stopping, intent-adaptive retrieval budgeting, and explicit cross-step context tracking. We released the anonymized logs, making them available at a public HuggingFace~\chref{https://huggingface.co/datasets/cx-cmu/deepresearchgym-agentic-search-logs}{repository}.

Neural retrieval models excel in Web search, but their training requires substantial amounts of labeled query-document pairs, which are costly to obtain. With the widespread availability of Web document collections like ClueWeb22, synthetic queries generated by large language models offer a scalable alternative. Still, synthetic training queries often vary in quality, which leads to suboptimal downstream retrieval performance. Existing methods typically filter out noisy query-document pairs based on signals from an external re-ranker. In contrast, we propose a framework that leverages Direct Preference Optimization (DPO) to integrate ranking signals into the query generation process, aiming to directly optimize the model towards generating high-quality queries that maximize downstream retrieval effectiveness. Experiments show higher ranker-assessed relevance between query-document pairs after DPO, leading to stronger downstream performance on the MS~MARCO benchmark when compared to baseline models trained with synthetic data.

The More Like This recommendation approach is ubiquitous in multiple domains and consists in recommending items similar to the one currently selected by the user, being particularly relevant when user data is scarce. We studied the impact of using semantic similarity in the context of the More Like This recommendation for mobile applications, by leveraging dense representations in order to infer the similarity between applications, based on their textual fields. Our approach was validated by comparing it to the solution currently in use by Aptoide, a mobile application store, since no benchmarks are available for this specific task. To further evaluate the proposed model, we asked 1262 users to compare the results achieved by both approaches, also allowing us to build an annotated dataset of similar applications. Results show that the semantic representations are able to capture the context of the applications, with more useful recommendations being presented to users, when compared to Aptoide’s current solution. For replication and future research, all the code and data used in this study was made publicly available, including two novel datasets (installed applications for more than one million users, and app user-labeled similarity), the fine-tuned model, and the test platform.

People often ask questions about places, and this is reflected on the frequency of geo-spatial queries made to information retrieval and question answering systems. Recent developments associated to these two types of systems rely on deep neural networks, specifically on methods for passage retrieval based on Transformer models, trained on large datasets like MS-MARCO. Despite significant progress in approaches for retrieving (or re-ranking) passages from a document collection according to their relevance to an input query, few studies have specifically looked at geo-spatial queries (i.e., where-questions directly concerning locations, and also questions covering other informational needs relating to places, their types, and affordances). In this work, we explore neural retrieval models in the context of geo-spatial queries, using a subset of MS-MARCO with questions and passages containing place-names. After characterizing the subset of MS-MARCO, we analyzed a re-ranking strategy based on geographic distance, which we argue to be useful for selecting hard negative examples for model training. Then, we fine-tuned neural ranking models, following bi-encoder or cross-encoder strategies, using the MS-MARCO subset together with a geographically-aware negative sampling procedure. Experimental results show that the fine-tuned models can indeed achieve a superior performance. We also describe a simple knowledge distillation procedure to further improve the computationally more efficient bi-encoder models, using the results of the cross-encoder.