Abstract Motivation Single-cell sequencing technologies allow researchers to study cell-cell variation within a cell population. Variations between cells are driven by the underlying biological network, particularly gene regulatory networks (GRNs). GRNs rewire as cells evolve, and different cells can have different GRNs. However, while single-cell RNA-sequencing (scRNA-seq) and single-cell multi-omics data have been used to reconstruct GRNs, the output GRNs are rarely cell-specific, but rather, most existing methods infer population-level or cell-type-level GRNs. Results We propose CeSpGRN (Cell-Specific Gene Regulatory Network inference), a method that infers cell-specific GRNs from scRNA-seq, paired scRNA-seq and scATAC-seq, or spatial transcriptomic data. In particular, existing methods that use matching scRNA-seq and scATAC-seq data incorporate population-level region information in GRN inference, whereas CeSpGRN utilizes single-cell resolution region information. CeSpGRN infers cell-specific GRNs using a kernel-weighted Gaussian Copula Graphical Model, and incorporates multi-omic or spatial location information when constructing the objective function. We tested CeSpGRN on both simulated and real datasets, and the results show that CeSpGRN has a superior performance compared to baseline methods in reconstructing GRNs and detecting regulatory interactions that differ between cells. CeSpGRN uncovered regulatory interactions that rewire during biological processes on real datasets. Availability and implementation CeSpGRN is a Python package available at https://github.com/PeterZZQ/CeSpGRN.

\noindent Session-based recommender systems capture the short-term interest of a user within a session. Session contexts (i.e., a user's high-level interests or intents within a session) are not explicitly given in most datasets, and implicitly inferring session context as an aggregation of item-level attributes is crude. In this paper, we propose \method, which implicitly contextualizes sessions. \method first generates implicit contexts for sessions by creating a session-item graph, learning graph embeddings, and clustering to assign sessions to contexts. \method then trains a session context predictor and uses the predicted contexts' embeddings to enhance the next-item prediction accuracy. Experiments on four datasets show that \method has superior next-item prediction accuracy than state-of-the-art models. A case study of \method on the Reddit dataset confirms that assigned session contexts are unique and meaningful.