Sequential recommendation has become indispensable in modern digital services. Prevalent recommendation techniques formulate the recommendation task with a language instruction fed into large language models (LLMs) to generate recommendations. However, the implicit interaction scenario of recommendation task cannot provide explicit reasoning supervision to activate LLM's multi-step reasoning capability. Besides, the manner of reasoning for enhancing recommendation is still underexplored. Therefore, we investigate activating multi-step reasoning with users' interactions and propose a multi-step reasoning-enhanced LLM (MSR-Rec), which tightly integrates reasoning with recommendation from designing reasoning chain to reasoning-based recommendation. A task-decomposed reasoning chain is elaborately designed to imitate users' thinking process, seamlessly involving reasoning into recommendation. Following the reasoning chain, MSR-Rec synthesizes reasoning supervision and fine-tunes LLM to adapt for task-specific reasoning. In inference, bidirectional reasoning is implemented from user and item sides, performing a closed-loop reasoning for recommendation. Comprehensive experiments demonstrate that MSR-Rec achieves the state-of-the-art performance in both recommendation quality and reasoning interpretability, advancing the integration of reasoning and recommendation in LLM-based systems.

Xinyu Zhang, Linmei Hu, Luhao Zhang, Wentao Cheng, Yashen Wang, Ge Shi, Chong Feng, Liqiang Nie. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

With the global epidemic of the COVID-19, various rumors spread wantonly on social networks, which has seriously affected the stability and harmony of the entire society. To purify the network environment, some researchers have proposed to fight rumors from the perspectives of tracing the source of rumors, detecting the authenticity of information, and predicting explosive fake news. But their works are fragmented, and their performance are not significant. So we need strong antirumor methods to fight rumors. To this end, this article proposes a more comprehensive antirumor mechanism, which can realize rumors source location, rumor detection, and popularity prediction (RLDP). In particular, in the task of localization, we propose graph neural network-based method, which does not need to specify the underlying propagation mode and the number of rumor sources; in the task of detection, utilizing lightGBM, we construct a rumor detection model; in the task of popularity prediction, we construct a model based on contrastive learning while considering user engagements and information propagation, and the text of rumor. Finally, we verify the performance of the proposed RLDP by conducting extensive experiments.

Nowadays, a large number of Knowledge Graph Completion (KGC) methods have been proposed by using embedding based manners, to overcome the incompleteness problem faced with knowledge graph (KG). One important recent innovation in Natural Language Processing (NLP) domain is the employ of deep neural models that make the most of pre-training, culminating in BERT, the most popular example of this line of approaches today. Recently, a series of new KGC methods introducing a pre-trained language model, such as KG-BERT, have been developed and released compelling performance. However, previous pre-training based KGC methods usually train the model by using simple training task and only utilize one-hop relational signals in KG, which leads that they cannot model high-order semantic contexts and multi-hop complex relatedness. To overcome this problem, this article presents a novel pre-training framework for KGC task, which especially consists of both one-hop relation level task (low-order) and multi-hop meta-graph level task (high-order). Hence, the proposed method can capture not only the elaborate sub-graph structure but also the subtle semantic information on the given KG. The empirical results show the efficiency of the proposed method on the widely used real-world datasets.

Despite substantial interest in applications of neural networks to information retrieval, neural ranking models have mostly been applied to conventional ad-hoc retrieval tasks over web pages and newswire articles. This article proposes a concept-enhanced pre-training model for microblog retrieval task, leveraging Semantic Matching Model (SMM) objective and Concept Correlation Model (CCM) objective. The proposed model is a novel neural ranking model specifically designed for ranking short-text microblog, which could merge the advantage of pre-training methodology for generating valid contextualized embedding with the superiority of the prior lexical knowledge (e.g., concept knowledge) for understanding short-text language semantic. We conduct experiments on widely used real-world datasets, and the experimental results demonstrate the efficiency of the proposed model, even compared with latest state-of-the-art neural-based models and pre-training based models.

Reasoning on knowledge graph (KG) has been studied for explainable recommendation due to its ability of providing explicit explanations. However, current KG-based explainable recommendation methods unfortunately ignore the temporal information (such as purchase time, recommend time, etc.), which may result in unsuitable explanations. In this work, we propose a novel Time-aware Path reasoning for Recommendation (TPRec for short) method, which leverages the potential of temporal information to offer better recommendation with plausible explanations. First, we present an efficient time-aware interaction relation extraction component to construct collaborative knowledge graph with time-aware interactions (TCKG for short), and then we introduce a novel time-aware path reasoning method for recommendation. We conduct extensive experiments on three real-world datasets. The results demonstrate that the proposed TPRec could successfully employ TCKG to achieve substantial gains and improve the quality of explainable recommendation.

Recent years have witnessed great advancement of representation learning (RL)-based models for the knowledge graph relation prediction task. However, they generally rely on structure information embedded in the encyclopedic knowledge graph, while the beneficial semantic information provided by lexical knowledge graph is ignored, leading the problem of shallow understanding and coarse-grained analysis for knowledge acquisition. Therefore, this article introduces concept information derived from the lexical knowledge graph (e.g., Probase), and proposes a novel Hierarchical Attention model for Relation Prediction, which consists of entity-level attention mechanism and concept-level attention mechanism, to throughly integrate multiple semantic signals. Experimental results demonstrate the efficiency of the proposed method on two benchmark datasets.

For guiding natural language generation, many semantic-driven methods have been proposed. While clearly improving the performance of the end-to-end training task, these existing semantic-driven methods still have clear limitations: for example, (i) they only utilize shallow semantic signals (e.g., from topic models) with only a single stochastic hidden layer in their data generation process, which suffer easily from noise (especially adapted for short-text etc.) and lack of interpretation; (ii) they ignore the sentence order and document context, as they treat each document as a bag of sentences, and fail to capture the long-distance dependencies and global semantic meaning of a document. To overcome these problems, we propose a novel semantic-driven language modeling framework, which is a method to learn a Hierarchical Language Model and a Recurrent Conceptualization-enhanced Gamma Belief Network, simultaneously. For scalable inference, we develop the auto-encoding Variational Recurrent Inference, allowing efficient end-to-end training and simultaneously capturing global semantics from a text corpus. Especially, this article introduces concept information derived from high-quality lexical knowledge graph Probase, which leverages strong interpretability and anti-nose capability for the proposed model. Moreover, the proposed model captures not only intra-sentence word dependencies, but also temporal transitions between sentences and inter-sentence concept dependence. Experiments conducted on several NLP tasks validate the superiority of the proposed approach, which could effectively infer meaningful hierarchical concept structure of document and hierarchical multi-scale structures of sequences, even compared with latest state-of-the-art Transformer-based models.

Since the length of microblog texts, such as tweets, is strictly limited to 140 characters, traditional Information Retrieval techniques suffer from the vocabulary mismatch problem severely and cannot yield good performance in the context of microblogosphere. To address this critical challenge, in this paper, we focus on the use of local conceptual word embeddings for enhance microblog retrieval effectiveness. In particular, we propose a novel k-Nearest Neighbor (kNN) based Query Expansion (QE) algorithm to generate words from local word embeddings to expand the original query, which leads to better understanding of the information need. Besides, in order to further satisfy users' real-time information need, we incorporate temporal evidences into the expansion algorithm, which can boost recent tweets in the retrieval results with respect to a given topic. Experimental results on the official TREC Twitter corpora demonstrate the significant superiority of our approach over baseline methods.

Person attribute recognition aims to identify the attribute labels from the pedestrian images. Extracting contextual relation from the images and attributes, including the spatial-semantic relations, the spatial context and the semantic correlation, is beneficial to enhance the discrimination of the features for recognizing the attributes. Thus, this work proposes a sequence contextual relation learning (SCRL) method to capture these relations. It first embeds the images and attributes into sequences in two branches. Then SCRL flexibly learns the contextual relation from the sequences with the parallel attention model structure, which integrates the inter-attention and intra-attention models. The inter-attention module is utilized to extract the spatial-semantic relations, while the intra-attention is designed to gain the spatial context and the semantic correlation. Both attention modules are comprised of several parallel attention units and each unit can obtain the pairwise relations in one subspace. Therefore, they obtain the relations in multiple subspaces, which can improve the comprehensiveness of the relation learning. Additionally, for the sake of better extraction of spatial-semantic relations, this paper employs connectionist temporal classification (CTC) loss which is capable of driving the network to enforce monotonic alignment between the image and attribute. It can also accelerate the convergence of the network by the algorithm in it. Extensive experiments on five public datasets, i.e., Market-1501 attribute, Duke attribute, PETA, RAP and PA-100K datasets, demonstrate the effectiveness of the proposed method.

Treatment effect plays an important role on decision making in many fields, such as social marketing, healthcare, and public policy. The key challenge on estimating treatment effect in the wild observational studies is to handle confounding bias induced by imbalance of the confounder distributions between treated and control units. Traditional methods remove confounding bias by re-weighting units with supposedly accurate propensity score estimation under the unconfoundedness assumption. Controlling high-dimensional variables may make the unconfoundedness assumption more plausible, but poses new challenge on accurate propensity score estimation. One strand of recent literature seeks to directly optimize weights to balance confounder distributions, bypassing propensity score estimation. But existing balancing methods fail to do selection and differentiation among the pool of a large number of potential confounders, leading to possible underperformance in many high-dimensional settings. In this article, we propose a data-driven Differentiated Confounder Balancing (DCB) algorithm to jointly select confounders, differentiate weights of confounders and balance confounder distributions for treatment effect estimation in the wild high-dimensional settings. Besides, under some settings with heavy confounding bias, in order to further reduce the bias and variance of estimated treatment effect, we propose a Regression Adjusted Differentiated Confounder Balancing (RA-DCB) algorithm based on our DCB algorithm by incorporating outcome regression adjustment. The synergistic learning algorithms we proposed are more capable of reducing the confounding bias in many observational studies. To validate the effectiveness of our DCB and RA-DCB algorithms, we conduct extensive experiments on both synthetic and real-world datasets. The experimental results clearly demonstrate that our algorithms outperform the state-of-the-art methods. By incorporating regression adjustment, our RA-DCB algorithm achieves more precise estimation on treatment effect than DCB algorithm, especially under the settings with heavy confounding bias. Moreover, we show that the top features ranked by our algorithm generate accurate prediction of online advertising effect.

This paper proposes a novel collaborative attention mechanism, to fully utilize the mutually reinforcing relationship among the knowledge graph representation learning procedure (i.e., structure representation) and textual relation representation learning procedure (i.e., text representation). Based on this collaborative attention mechanism, a text-enhanced knowledge graph (KG) representation model is proposed, which could utilize textual information to enhance the knowledge representations and make the multi-direction signals to be fully integrated to learn more accurate textual representations for further improving structure representation and vice versa. Experimental results demonstrate the efficiency of the proposed model on both link prediction task and triple classification task.

Most short-text embedding models typically represent each short-text only using the literal meanings of the words, which makes these models indiscriminative for the ubiquitous polysemy. In order to enhance the semantic representation capability of the short-texts, we (i) propose a novel short-text conceptualization algorithm to assign the associated concepts for each short-text, and then (ii) introduce the conceptualization results into learning the conceptual short-text embeddings. Hence, this semantic representation is more expressive than some widely-used text representation models such as the latent topic model. Wherein, the short-text conceptualization algorithm used here is based on a novel co-ranking framework, enabling the signals (i.e., the words and the concepts) to fully interplay to derive the solid conceptualization for the short-texts. Afterwards, we further extend the conceptual short-text embedding models by utilizing an attention-based model that selects the relevant words within the context to make more efficient prediction. The experiments on the real-world datasets demonstrate that the proposed conceptual short-text embedding model and short-text conceptualization algorithm are more effective than the state-of-the-art methods.

We tackle the problem of improving microblog retrieval algorithms by proposing a Feedback Concept Model for query expansion. In particular, we expand the query using knowledge information derived from Probase so that the expanded one could better reflect users' search intent, which allows for microblog retrieval at a concept-level, rather than term-level. In the proposed feedback concept model: (i) we mine the concept information implicit in short-texts based on the external knowledge bases; (ii) with the relevant concepts associated with short-texts, a mixture model is generated to estimate a concept language model; (iii) finally, we utilize the concept language model for query expansion. Moreover, we incorporate temporal prior into the proposed query expansion method to satisfy real-time information need. Finally, we test the generalization power of the feedback concept model on the TREC Microblog corpora. The experimental results demonstrate that the proposed model outperforms the previous methods for microblog retrieval significantly.

Most sentence embedding models typically represent each sentence only using word surface, which makes these models indiscriminative for ubiquitous homonymy and polysemy.In order to enhance representation capability of sentence, we employ conceptualization model to assign associated concepts for each sentence in the text corpus, and then learn conceptual sentence embedding (CSE).Hence, this semantic representation is more expressive than some widely-used text representation models such as latent topic model, especially for short-text.Moreover, we further extend CSE models by utilizing a local attention-based model that select relevant words within the context to make more efficient prediction.In the experiments, we evaluate the CSE models on two tasks, text classification and information retrieval.The experimental results show that the proposed models outperform typical sentence embed-ding models.

Accurate road speed predictions can help drivers in smart route planning. Although the issue has been studied previously, most existing work focus on arterial roads only, where sensors are configured closely for collecting complete real-time data. For collector roads where sensors sparsly cover, however, speed predictions are often ignored. With GPS-equipped floating car signals being available nowadays, we aim at forecasting collector road speeds by utilizing these signals. The main challenge compared with arterial roads comes from the missing data. In a time slot of the real case, over 90% of collector roads cannot be covered by enough floating cars. Thus most traditional approaches for arterial roads, relying on complete historical data, cannot be employed directly. Aiming at solving this problem, we propose a multi-view road speed prediction framework. In the first view, temporal patterns are modeled by a layered hidden Markov model; and in the second view, spatial patterns are modeled by a collective matrix factorization model. The two models are learned and inferred simultaneously in a co-regularized manner. Experiments conducted in the Beijing road network, based on 10K taxi signals in 2 years, have demonstrated that the approach outperforms traditional approaches by 10% in MAE and RMSE.