The integration of Large Language Models (LLMs) with retrieval systems has shown promising potential in retrieving documents (docs) or advertisements (ads) for a given query.Existing LLM-based retrieval methods generate numeric or content-based DocIDs to retrieve docs or ads.However, the one-to-few mapping between numeric IDs and docs, along with the time-consuming content extraction, leads to semantic inefficiency and limits the scalability of existing methods on large-scale corpora.In this paper, we propose the Realtime Ad REtrieval (RARE) framework, which leverages LLM-generated text called Commercial Intentions (CIs) as an intermediate semantic representation to directly retrieve ads for queries in real-time.These CIs are generated by a customized LLM injected with commercial knowledge, enhancing its domain relevance.Each CI corresponds to multiple ads, yielding a lightweight and scalable set of CIs.RARE has been implemented in a real-world online system, handling daily search volumes in billions.The online implementation has yielded significant benefits: a 5.04% increase in consumption, a 6.37% increase in gross merchandise volume (GMV), a 1.28% enhancement in click-through rate (CTR) and a 5.29% increase in shallow conversions.Extensive offline experiments show RARE's superiority over ten competitive baselines in four major categories.

It is crucial for the photovoltaic system to have an accurate model and well‐estimated parameters to further increase conversion efficiency. Most existing methods for identifying photovoltaic model parameters have problems such as high computational cost, local optimum trouble, or difficulty in providing the best performance due to complex adjustments of algorithm parameters. To improve these defects, a hunter‐prey optimization algorithm coordinating mutual benefit and sharing and interactive learning activities (EHPO) is proposed. First, based on hunter‐prey optimization, timely information sharing with the mutual benefit and sharing mechanism was achieved when the algorithm was applied to search for prey, thus improving the searching precision and convergence rate of the algorithm. Second, hunters used the history optimization information and useful information from peers to guide the search direction, thus balancing the global and local development abilities of the algorithm. Finally, a lens imaging backward learning strategy was adopted to prevent the algorithm from falling into the local optimum, thus increasing the diversity of varieties and the probability of finding the global optimal solution. The simulation results of the single‐diode model (SDM), double‐diode model (DDM), triple‐diode model (TDM), and other PV models in different environmental conditions show that the improved EHPO algorithm is more advantageous for parameter extraction than other advanced metaheuristic algorithms. This study demonstrates that EHPO is an accurate and reliable tool for predicting the parameters of PV models.