Near‐infrared (NIR) phosphor‐converted light‐emitting diodes (pc‐LEDs) hold great potential for applications ranging from night vision to non‐destructive detection. However, it remains a long‐standing challenge to develop NIR phosphors simultaneously with longer‐wavelength broadband emissions and higher efficiency. Herein, ultra‐broadband Ga4GeO8:Cr3+ (GGO:Cr3+) phosphors are developed, with the NIR emission covering 700–1300 nm. Furthermore, tunable emission bands peaking from 835 to 980 nm are achieved simply by varying the Cr3+ concentration. Particularly, emission maxima (λmax) of GGO:xCr3+ shift from 850 to 900 nm without intensity loss when increasing x values between 0.02 and 0.10. An internal quantum yield of 60% is achieved for GGO:0.02Cr3+ (λmax ≈ 850 nm, full width at half maximum (FWHM) ≈215 nm). The origin of tunable ultra‐broadband emissions of GGO:Cr3+ is revealed on the basis of structural and time‐resolved spectroscopic analysis. The pc‐LED fabricated by GGO:0.02Cr3+ exhibits a maximum NIR output power of ≈56 mW at 400 mA drive current, and its application in high‐penetration quality analysis of fruits is also demonstrated. The results indicate that GGO:Cr3+ phosphors have high promise for practical applications in NIR pc‐LED devices.

Recently, collaborative filtering combined with various kinds of deep learning models is appealing to recommender systems, which have shown a strong positive effect in an accuracy improvement. However, many studies related to deep learning model rely heavily on abundant information to improve prediction accuracy, which has stringent data requirements in addition to raw rating data. Furthermore, most of them ignore the interaction effect between users and items when building the recommendation model. To address these issues, we propose DCCR, a deep collaborative conjunctive recommender, for rating prediction tasks that are solely based on the raw ratings. A DCCR is a hybrid architecture that consists of two different kinds of neural network models (i.e., an autoencoder and a multilayered perceptron). The main function of the autoencoder is to extract the latent features from the perspectives of users and items in parallel, while the multilayered perceptron is used to represent the interaction between users and items based on fusing the user and item latent features. To further improve the performance of DCCR, an advanced activation function is proposed, which can be specified with input vectors. The extensive experiments conducted with two well-known real-world datasets and performances of the DCCR with varying settings are analyzed. The results demonstrate that our DCCR model outperforms other state-of-art methods. We also discuss the performance of the DCCR with additional layers to show the extensibility of our model. INDEX TERMS Recommender systems, collaborative filtering, rating prediction, denoising autoencoders, multi layered perceptron.

No abstract