Deep learning provides appropriate mechanisms to predict vessel trajectories for safer and efficient shipping, but still existing models are mainly oriented to longer-term prediction trends and do not fully support real time navigation needs. While most recent works have been largely exploiting Automatic Identification System (AIS), the complete semantics of these data haven't so far fully exploited. The research presented in this paper introduced an extended sequence-to-sequence model using AIS data. A Gated Recurrent Unit (GRU) network encodes historical spatio-temporal sequences as a context vector, which not only preserves the sequential relationships among trajectory locations, but also alleviates the gradient descent problem. The GRU network acts as a decoder, outputting target trajectory location sequences. Real AIS data from the Chongqing and Wuhan sections of the Yangzi River were selected as typical experimental areas for evaluation purposes. The proposed ST-Seq2Seq model has been tested against the LSTM-RNN and GRU-RNN baseline models for short term trajectory prediction experiments. A 10-minute historical trajectory sequence was used to predict the trajectory sequence for the next five minutes. Overall, the findings show that LSTM and GRU networks, while applying a recursive method to predict a sequence of continuous trajectory points, when the number of predicted trajectory points increases accuracy decreases. Conversely, the extended sequence-to-sequence model shows satisfactory stability on different ship channels.

Although matrix model-based approaches to collaborative filtering (CF), such as latent factor models, achieve good accuracy in review rating prediction, they still face data sparsity problems. Many recent studies have exploited review text information to improve the performance of predictions. The review content that they use, however, is usually on the coarse-grained text level or sentence level. In this paper, we propose a joint model that incorporates review text information with matrix factorization for review rating prediction. First, we adopt an aspect extraction method and propose a simple and practical algorithm to represent the review by aspects and sentiments. Then, we propose two similarity measures: aspect-based user similarity and aspect-based product similarity. Finally, aspect-based user and product similarity measures are incorporated into a matrix factorization to build a joint model for rating prediction. To this end, our model can alleviate the data sparsity problem and obtain interpretability for the recommendation. We conducted experiments on two datasets. The experimental results demonstrate the effectiveness of the proposed model.

An increasing number of people are choosing to shop online; hence, online reviews are an increasingly influential factor in consumer purchasing decisions. However, extracting useful information from online reviews is a challenge in the analysis of consumer sentiment. In this paper, we focus on the automatic discovery of the features evaluated in online reviews and the expression of sentiment. We propose a novel fine-grained topic model called the “review unit topic model” (RUTM) to extract semantic meanings and polarities. In this model, a review unit rather than a review sentence is treated as the representational model, and prior knowledge of sentiment is further exploited to identify aspect-aware sentiment polarities. We evaluate RUTM extensively using real-world review data. Experimental results demonstrate that the proposed model outperforms well-established baseline models in sentiment analysis tasks.

Aiming at the problem that the GitHub platform only supports the retrieval of developers through their usernames and it is difficult to directly obtain developers' expertise information, this paper proposes an open source domain expert retrieval model (OSDERM) based on the network representation learning algorithm OSC2vec (Open Source Collaboration to Vector). The model mainly consists of two core parts: Expert Profiling and Expert Finding. Expert Profiling aims to enrich the expertise information in the search results by labeling the expertise of developers; while Expert Finding achieves rapid location of the most suitable domain experts through keyword matching, which greatly saves the time and effort of searching for experts in the open source community. Experiments using the GitHub ecological dataset show that the model outperforms existing comparative algorithms in discovering open source domain experts, and can provide an effective reference for enterprise recruitment