Recent advances in large language models have led to the widespread integration of chatbots as built-in features in modern applications. To model user latent intent and provide topic guidance, the question recommendation task serves as the entry point for built-in chatbots. In content-consuming applications, interaction data with built-in chatbots exhibit notable sparsity, while consuming signals account for the majority of user behaviors. Therefore, it is reasonable to incorporate content-consuming signals for user preference modeling. However, a significant domain shift exists between the preference in the content-consuming domain and that in the conversational topic domain. It is non-trivial to transfer those cross-domain signals into meaningful suggested topics that are aligned with users' real-time preference.

Large Language Model (LLM)-driven conversational search is shifting information retrieval from reactive keyword matching to proactive, open-ended dialogues. In this context, Conversation Starters are widely deployed to provide personalized query recommendations that help users initiate dialogues. Conventionally, recommending these starters relies on a closed ''exposure-click'' loop. Yet, this feedback loop mechanism traps the system in an echo chamber where, compounded by data sparsity, it fails to capture the dynamic nature of conversational search intents shaped by the open world. As a result, the system skews towards popular but generic suggestions. In this work, we uncover an untapped paradigm shift to shatter this harmful feedback loop: harnessing user ''free will'' through active user expressions. Unlike traditional recommendations, conversational search empowers users to bypass menus entirely through manually typed queries. The open-world intents in active queries hold the key to breaking this loop. However, incorporating them is non-trivial: (1) there exists an inherent distribution shift between active queries and formulated starters. (2) Furthermore, the ''non-ID-able'' nature of open text renders traditional item-based popularity statistics ineffective for large-scale industrial streaming training. To this end, we propose Passive-Active Bridge (PA-Bridge), a novel framework that employs an adversarial distribution aligner to bridge the distributional gap between passively recommended starters and active expressions. Moreover, we introduce a semantic discretizer to enable the deployment of popularity debiasing algorithms. Online A/B tests on our platform, which serves hundreds of millions of users, demonstrate that PA-Bridge significantly boosts the Feature Penetration Rate by 0.54% and User Active Days by 0.04%.

Recently, motivated by the outstanding achievements of diffusion models, the diffusion process has been employed to strengthen representation learning in recommendation systems. Most diffusion-based recommendation models typically utilize standard Gaussian noise in symmetric forward and reverse processes in continuous data space. Nevertheless, the samples derived from recommendation systems inhabit a discrete data space, which is fundamentally different from the continuous one. Moreover, Gaussian noise has the potential to corrupt personalized information within latent representations. In this work, we propose a novel and effective method, named Asymmetric Diffusion Recommendation Model (AsymDiffRec), which learns forward and reverse processes in an asymmetric manner. We define a generalized forward process that simulates the missing features in real-world recommendation samples. The reverse process is then performed in an asymmetric latent feature space. To preserve personalized information within the latent representation, a task-oriented optimization strategy is introduced. In the serving stage, the raw sample with missing features is regarded as a noisy input to generate a denoising and robust representation for the final prediction. By equipping base models with AsymDiffRec, we conduct online A/B tests, achieving improvements of +0.131% and +0.166% in terms of users' active days and app usage duration respectively. Additionally, the extended offline experiments also demonstrate improvements. AsymDiffRec has been implemented in the Douyin Music App.

User preferences follow a dynamic pattern over a day, e.g., at 8 am, a user might prefer to read news, while at 8 pm, they might prefer to watch movies. Time modeling aims to enable recommendation systems to perceive time changes to capture users' dynamic preferences over time, which is an important and challenging problem in recommendation systems. Especially, streaming recommendation systems in the industry, with only available samples of the current moment, present greater challenges for time modeling. There is still a lack of effective time modeling methods for streaming recommendation systems. In this paper, we propose an effective and universal method Interest Clock to perceive time information in recommendation systems. Interest Clock first encodes users' time-aware preferences into a clock (hour-level personalized features) and then uses Gaussian distribution to smooth and aggregate them into the final interest clock embedding according to the current time for the final prediction. By arming base models with Interest Clock, we conduct online A/B tests, obtaining +0.509% and +0.758% improvements on user active days and app duration respectively. Besides, the extended offline experiments show improvements as well. Interest Clock has been deployed on Douyin Music App.

In article recommendation, the amount of time user spends on viewing articles, dwell time, is an important metric to measure the post-click engagement of user on content and has been widely used as a proxy to user satisfaction, complementing the click feedback. Recently, the sequential pattern of impression-click-read has become one of the most popular type of article recommendation service in real world, where users are presented with a list of titles at first, then get interested in one and click in for reading. Predicting dwell time in such service is conditioned on the click, since the user reads the article only after he clicks the corresponding title. We argue that conventional models for dwell time prediction, which mainly focus on the relevance between the content and the general preference of user, are not well-designed for such service. There is a natural assumption in recommendation system that the click indicates user's getting attracted by the item. Therefore, in the pattern of impression-click-read, the user might get interested and curious on some other concepts different from his general preference while reading, due to the attraction of the title. Conventional models tend to ignore the gap between such temporary interest and the general preference of user in the reading behavior, which fails to use the pattern of impression-click-read and the assumption of the click very well. In this work, we propose a framework, Click-guide Network (CGN) for dwell time prediction, which makes good use of the sequential pattern and the assumption to model the ”guidance” of the click on user preference. CGN is a joint learner for dwell time and click through rate (CTR). We introduce the CTR task as an auxiliary task to help us better learn the preference of user and the representation of title. Besides, we propose the Guider to capture the user's temporary interest raised by the title. We collect the data from WeChat, a widely-used mobile app in China, for experiments. The results demonstrate the advantages of CGN over several competitive baselines on dwell time prediction, while our case studies show how the Guider effectively capture the temporary interest of user.

Latent Factor Models (LFMs) based on Collaborative Filtering (CF) have been widely applied in many recommendation systems, due to their good performance of prediction accuracy. In addition to users' ratings, auxiliary information such as item features is often used to improve performance, especially when ratings are very sparse. To the best of our knowledge, most existing LFMs integrate different item features in the same way for all users. Nevertheless, the attention on different item attributes varies a lot from user to user. For personalized recommendation, it is valuable to know what feature of an item a user cares most about. Besides, the latent vectors used to represent users or items in LFMs have few explicit meanings, which makes it difficult to explain why an item is recommended to a specific user. In this work, we propose the Attention-driven Factor Model (AFM), which can not only integrate item features driven by users' attention but also help answer this "why". To estimate users' attention distributions on different item features, we propose the Gated Attention Units (GAUs) for AFM. The GAUs make it possible to let the latent factors "talk", by generating user attention distributions from user latent vectors. With users' attention distributions, we can tune the weights of item features for different users. Moreover, users' attention distributions can also serve as explanations for our recommendations. Experiments on several real-world datasets demonstrate the advantages of AFM (using GAUs) over competitive baseline algorithms on rating prediction.