arXiv subjects

Zhichao Xu

Publications and source records attributed to Zhichao Xu.

61 records · Page 4Linked to original sources

arXiv2021

Understanding the Effectiveness of Reviews in E-commerce Top-N Recommendation

Modern E-commerce websites contain heterogeneous sources of information, such as numerical ratings, textual reviews and images. These information can be utilized to assist recommendation. Through textual reviews, a user explicitly express her affinity towards the item. Previous researchers found that by using the information extracted from these reviews, we can better profile the users' explicit preferences as well as the item features, leading to the improvement of recommendation performance. However, most of the previous algorithms were only utilizing the review information for explicit-feedback problem i.e. rating prediction, and when it comes to implicit-feedback ranking problem such as top-N recommendation, the usage of review information has not been fully explored. Seeing this gap, in this work, we investigate the effectiveness of textual review information for top-N recommendation under E-commerce settings. We adapt several SOTA review-based rating prediction models for top-N recommendation tasks and compare them to existing top-N recommendation models from both performance and efficiency. We find that models utilizing only review information can not achieve better performances than vanilla implicit-feedback matrix factorization method. When utilizing review information as a regularizer or auxiliary information, the performance of implicit-feedback matrix factorization method can be further improved. However, the optimal model structure to utilize textual reviews for E-commerce top-N recommendation is yet to be determined.

cs.IR↗

arXiv2021

A Zero Attentive Relevance Matching Networkfor Review Modeling in Recommendation System

User and item reviews are valuable for the construction of recommender systems. In general, existing review-based methods for recommendation can be broadly categorized into two groups: the siamese models that build static user and item representations from their reviews respectively, and the interaction-based models that encode user and item dynamically according to the similarity or relationships of their reviews. Although the interaction-based models have more model capacity and fit human purchasing behavior better, several problematic model designs and assumptions of the existing interaction-based models lead to its suboptimal performance compared to existing siamese models. In this paper, we identify three problems of the existing interaction-based recommendation models and propose a couple of solutions as well as a new interaction-based model to incorporate review data for rating prediction. Our model implements a relevance matching model with regularized training losses to discover user relevant information from long item reviews, and it also adapts a zero attention strategy to dynamically balance the item-dependent and item-independent information extracted from user reviews. Empirical experiments and case studies on Amazon Product Benchmark datasets show that our model can extract effective and interpretable user/item representations from their reviews and outperforms multiple types of state-of-the-art review-based recommendation models.

cs.IR↗

arXiv2021

Using Sampling Strategy to Assist Consensus Sequence Analysis

Consensus Sequences of event logs are often used in process mining to quickly grasp the core sequence of events to be performed in a process, or to represent the backbone of the process for doing other analyses. However, it is still not clear how many traces are enough to properly represent the underlying process. In this paper, we propose a novel sampling strategy to determine the number of traces necessary to produce a representative consensus sequence. We show how to estimate the difference between the predefined Expert Model and the real processes carried out. This difference level can be used as reference for domain experts to adjust the Expert Model. In addition, we apply this strategy to several real-world workflow activity datasets as a case study. We show a sample curve fitting task to help readers better understand our proposed methodology.

cs.AI↗

arXiv2020

E-commerce Recommendation with Weighted Expected Utility

Different from shopping at retail stores, consumers on e-commerce platforms usually cannot touch or try products before purchasing, which means that they have to make decisions when they are uncertain about the outcome (e.g., satisfaction level) of purchasing a product. To study people's preferences, economics researchers have proposed the hypothesis of Expected Utility (EU) that models the subject value associated with an individual's choice as the statistical expectations of that individual's valuations of the outcomes of this choice. Despite its success in studies of game theory and decision theory, the effectiveness of EU, however, is mostly unknown in e-commerce recommendation systems. Previous research on e-commerce recommendation interprets the utility of purchase decisions either as a function of the consumed quantity of the product or as the gain of sellers/buyers in the monetary sense. As most consumers just purchase one unit of a product at a time and most alternatives have similar prices, such modeling of purchase utility is likely to be inaccurate in practice. In this paper, we interpret purchase utility as the satisfaction level a consumer gets from a product and propose a recommendation framework using EU to model consumers' behavioral patterns. We assume that consumer estimates the expected utilities of all the alternatives and choose products with maximum expected utility for each purchase. To deal with the potential psychological biases of each consumer, we introduce the usage of Probability Weight Function (PWF) and design our algorithm based on Weighted Expected Utility (WEU). Empirical study on real-world e-commerce datasets shows that our proposed ranking-based recommendation framework achieves statistically significant improvement against both classical Collaborative Filtering/Latent Factor Models and state-of-the-art deep models in top-K recommendation.

cs.IR↗

arXiv2014

Dual-wavelength active optical clock

We experimentally realize the dual-wavelength active optical clock for the first time. As the Cs cell temperature is kept between 118 $^{\circ }C$ and 144 $^{\circ }C$, both the 1359 nm and the 1470 nm stimulated emission output of Cs four-level active optical clock are detected. The 1470 nm output linewidth of each experimental setup of Cs four-level active optical clock is measured to be 590 Hz with the main cavity length unstabilized. To stabilize the cavity length of active optical clock, the experimental scheme of 633 nm and 1359 nm good-bad cavity dual-wavelength active optical clock is proposed, where 633 nm and 1359 nm stimulated emission is working at good-cavity and bad-cavity regime respectively. The cavity length is stabilized by locking the 633 nm output frequency to a super-cavity with the Pound-Drever-Hall (PDH) technique. The frequency stability of 1359 nm bad-cavity stimulated emission output is then expected to be further improved by at least 1 order of magnitude than the 633 nm PDH system due to the suppressed cavity pulling effect of active optical clock, and the quantum limited linewidth of 1359 nm output is estimated to be 77.6 mHz.

physics.atom-ph↗

arXiv2014

Towards Efficient Path Query on Social Network with Hybrid RDF Management

The scalability and exibility of Resource Description Framework(RDF) model make it ideally suited for representing online social networks(OSN). One basic operation in OSN is to find chains of relations,such as k-Hop friends. Property path query in SPARQL can express this type of operation, but its implementation suffers from performance problem considering the ever growing data size and complexity of OSN.In this paper, we present a main memory/disk based hybrid RDF data management framework for efficient property path query. In this hybrid framework, we realize an efficient in-memory algebra operator for property path query using graph traversal, and estimate the cost of this operator to cooperate with existing cost-based optimization. Experiments on benchmark and real dataset demonstrated that our approach can achieve a good tradeoff between data load expense and online query performance.

cs.DB↗

arXiv2014

Lasing and suppressed cavity-pulling effect of Cesium active optical clock

We experimentally demonstrate the collective emission behavior and suppressed cavity-pulling effect of four-level active optical clock with Cesium atoms. Thermal Cesium atoms in a glass cell velocity selective pumped with a 455.5 nm laser operating at 6S$_{1/2}$ to 7P$_{3/2}$ transition are used as lasing medium. Population inverted Cesium atoms between 7S$_{1/2}$ and 6P$_{3/2}$ levels are optical weakly coupled by a pair cavity mirrors working at deep bad-cavity regime with a finesse of 4.3, and the ratio between cavity bandwidth and gain bandwidth is approximately 45. With increased 455.5 nm pumping laser intensity, the output power of cesium active optical clock at 1469.9 nm from 7S$_{1/2}$ level to 6P$_{3/2}$ level shows a threshold and reach a power of 13 $μ$W. Active optical clock would dramatically improve the optical clock stability since the lasing frequency does not follow the cavity length variation exactly, but in a form of suppressed cavity pulling effect. In this letter the cavity pulling effect is measured using a Fabry-Perot interferometer (FPI) to be reduced by a factor of 38.2 and 41.4 as the detuning between the 1469.9 nm cavity length of the Cs active optical clock and the Cs 1469.9 nm transition is set to be 140.8 MHz and 281.6 MHz respectively. The mechanism demonstrated here is of great significance for new generation optical clocks and can be applied to improve the stability of best optical clocks by at least two orders of magnitude.

physics.atom-ph↗