Search for: All records

Explanations in a recommender system assist users make informed decisions among a set of recommended items. Extensive research attention has been devoted to generate natural language explanations to depict how the recommendations are generated and why the users should pay attention to them. However, due to different limitations of those solutions, e.g., template-based or generation-based, it is hard to make the explanations easily perceivable, reliable, and personalized at the same time. In this work, we develop a graph attentive neural network model that seamlessly integrates user, item, attributes and sentences for extraction-based explanation. The attributes of items are selected as the intermediary to facilitate message passing for user-item specific evaluation of sentence relevance. And to balance individual sentence relevance, overall attribute coverage and content redundancy, we solve an integer linear programming problem to make the final selection of sentences. Extensive empirical evaluations against a set of state-of-the-art baseline methods on two benchmark review datasets demonstrated the generation quality of proposed solution. 

As recommendation is essentially a comparative (or ranking) process, a good explanation should illustrate to users why an item is believed to be better than another, i.e., comparative explanations about the recommended items. Ideally, after reading the explanations, a user should reach the same ranking of items as the system’s. Unfortunately, little research attention has yet been paid on such comparative explanations. In this work, we develop an extract-and-refine architecture to explain the relative comparisons among a set of ranked items from a recommender system. For each recommended item, we first extract one sentence from its associated reviews that best suits the desired comparison against a set of reference items. Then this extracted sentence is further articulated with respect to the target user through a generative model to better explain why the item is recommended. We design a new explanation quality metric based on BLEU to guide the end-to-end training of the extraction and refinement components, which avoids generation of generic content. Extensive offline evaluations on two large recommendation benchmark datasets and serious user studies against an array of state-of-the-art explainable recommendation algorithms demonstrate the necessity of comparative explanations and the effectiveness of our solution. 

Sequential recommendation is the task of predicting the next items for users based on their interaction history. Modeling the dependence of the next action on the past actions accurately is crucial to this problem. Moreover, sequential recommendation often faces serious sparsity of item-to-item transitions in a user's action sequence, which limits the practical utility of such solutions. To tackle these challenges, we propose a Category-aware Collaborative Sequential Recommender. Our preliminary statistical tests demonstrate that the in-category item-to-item transitions are often much stronger indicators of the next items than the general item-to-item transitions observed in the original sequence. Our method makes use of item category in two ways. First, the recommender utilizes item category to organize a user's own actions to enhance dependency modeling based on her own past actions. It utilizes self-attention to capture in-category transition patterns, and determines which of the in-category transition patterns to consider based on the categories of recent actions. Second, the recommender utilizes the item category to retrieve users with similar in-category preferences to enhance collaborative learning across users, and thus conquer sparsity. It utilizes attention to incorporate in-category transition patterns from the retrieved users for the target user. Extensive experiments on two large datasets prove the effectiveness of our solution against an extensive list of state-of-the-art sequential recommendation models. 

Textual information, such as news articles, social media, and online forum discussions, often comes in a form of sequential text streams. Events happening in the real world trigger a set of articles talking about them or related events over a period of time. In the meanwhile, even one event is fading out, another related event could raise public attention. Hence, it is important to leverage the information about how topics influence each other over time to obtain a better understanding and modeling of document streams. In this paper, we explicitly model mutual influence among topics over time, with the purpose to better understand how events emerge, fade and inherit. We propose a temporal point process model, referred to as Correlated Temporal Topic Model (CoTT), to capture the temporal dynamics in a latent topic space. Our model allows for efficient online inference, scaling to continuous time document streams. Extensive experiments on real-world data reveal the effectiveness of our model in recovering meaningful temporal dependency structure among topics and documents. 

Modern buildings produce thousands of data streams, and the ability to automatically infer the physical context of such data is the key to enabling building analytics at scale. As acquiring this contextual information is currently a time-consuming and error-prone manual process, in this study we make the first attempt at automatically inferring one important contextual aspect of the equipment in buildings --- how each equipment is functionally connected with another. The main insight behind our solution is that functionally connected equipment is exposed to the same events in the physical world, creating correlated changes in the time series data of both equipment. Because events are of indeterminate length in time series, however, identifying them requires solving a non-polynomial combinatorial data segmentation problem. We present a solution that first extracts latent events from the sensory time series data, and then sifts out coincident events with a customized correlation procedure to identify the relationship between equipment. We evaluated our approach on data collected from over 1,000 pieces of equipment from 5 commercial buildings of various sizes located in different geographical regions in the US. Results show that this approach achieves 94.38% accuracy in relation inference, compared to 85.49% by the best baseline. 

The massively available data about user engagement with online information service systems provides a gold mine about users' latent intents. It calls for quantitative user behavior modeling. In this paper, we study the problem by looking into users' sequential interactive behaviors. Inspired by the concepts of episodic memory and semantic memory in cognitive psychology, which describe how users' behaviors are differently influenced by past experience, we propose a Long- and Short-term Hawkes Process model. It models the short-term dependency between users' actions within a period of time via a multi-dimensional Hawkes process and the long-term dependency between actions across different periods of time via a one dimensional Hawkes process. Experiments on two real-world user activity log datasets (one from an e-commerce website and one from a MOOC website) demonstrate the effectiveness of our model in capturing the temporal dependency between actions in a sequence of user behaviors. It directly leads to improved accuracy in predicting the type and the time of the next action. Interestingly, the inferred dependency between actions in a sequence sheds light on the underlying user intent behind direct observations and provides insights for downstream applications.