LegalGNN: Legal Information Enhanced Graph Neural Network for Recommendation

2022 ◽  
Vol 40 (2) ◽  
pp. 1-29
Author(s):  
Jun Yang ◽  
Weizhi Ma ◽  
Min Zhang ◽  
Xin Zhou ◽  
Yiqun Liu ◽  
...  

Recommendation in legal scenario (Legal-Rec) is a specialized recommendation task that aims to provide potential helpful legal documents for users. While there are mainly three differences compared with traditional recommendation: (1) Both the structural connections and textual contents of legal information are important in the Legal-Rec scenario, which means feature fusion is very important here. (2) Legal-Rec users prefer the newest legal cases (the latest legal interpretation and legal practice), which leads to a severe new-item problem. (3) Different from users in other scenarios, most Legal-Rec users are expert and domain-related users. They often concentrate on several topics and have more stable information needs. So it is important to accurately model user interests here. To the best of our knowledge, existing recommendation work cannot handle these challenges simultaneously. To address these challenges, we propose a legal information enhanced graph neural network–based recommendation framework (LegalGNN). First, a unified legal content and structure representation model is designed for feature fusion, where the Heterogeneous Legal Information Network (HLIN) is constructed to connect the structural features (e.g., knowledge graph) and contextual features (e.g., the content of legal documents) for training. Second, to model user interests, we incorporate the queries users issued in legal systems into the HLIN and link them with both retrieved documents and inquired users. This extra information is not only helpful for estimating user preferences, but also valuable for cold users/items (with less interaction history) in this scenario. Third, a graph neural network with relational attention mechanism is applied to make use of high-order connections in HLIN for Legal-Rec. Experimental results on a real-world legal dataset verify that LegalGNN outperforms several state-of-the-art methods significantly. As far as we know, LegalGNN is the first graph neural model for legal recommendation.

2017 ◽  
Vol 10 (1) ◽  
pp. 77
Author(s):  
Adnan Imeri ◽  
Abdelaziz Khadraoui ◽  
Thang Le Dinh ◽  
Djamel Khadraoui

This paper aims at presenting a theoretical approach for the representation of legal and ethical information in a personalized way in ICT platforms. The personalization process allows us to adapt the legal information for end- users of ICT platforms. This approach points out the key element for defining a complete legal and ethical corpus of documents. Based on the corresponding scenario, the legal information is selected, filtered and then presented in ICT platforms. For organizating and analyzing this legal and ethical information, we reuse the existing notion of the decision tree and enhance it with a set of legal parameters emerging from legal documents. This legal decision tree will help us to treat different types of legal cases, which may relate to different service scenarios.


Author(s):  
A. Syahputra

Surveillance is very important in managing a steamflood project. On the current surveillance plan, Temperature and steam ID logs are acquired on observation wells at least every year while CO log (oil saturation log or SO log) every 3 years. Based on those surveillance logs, a dynamic full field reservoir model is updated quarterly. Typically, a high depletion rate happens in a new steamflood area as a function of drainage activities and steamflood injection. Due to different acquisition time, there is a possibility of misalignment or information gaps between remaining oil maps (ie: net pay, average oil saturation or hydrocarbon pore thickness map) with steam chest map, for example a case of high remaining oil on high steam saturation interval. The methodology that is used to predict oil saturation log is neural network. In this neural network method, open hole observation wells logs (static reservoir log) such as vshale, porosity, water saturation effective, and pay non pay interval), dynamic reservoir logs as temperature, steam saturation, oil saturation, and acquisition time are used as input. A study case of a new steamflood area with 16 patterns of single reservoir target used 6 active observation wells and 15 complete logs sets (temperature, steam ID, and CO log), 19 incomplete logs sets (only temperature and steam ID) since 2014 to 2019. Those data were divided as follows ~80% of completed log set data for neural network training model and ~20% of completed log set data for testing the model. As the result of neural model testing, R2 is score 0.86 with RMS 5% oil saturation. In this testing step, oil saturation log prediction is compared to actual data. Only minor data that shows different oil saturation value and overall shape of oil saturation logs are match. This neural network model is then used for oil saturation log prediction in 19 incomplete log set. The oil saturation log prediction method can fill the gap of data to better describe the depletion process in a new steamflood area. This method also helps to align steam map and remaining oil to support reservoir management in a steamflood project.


2020 ◽  
Vol 54 (1) ◽  
pp. 1-12
Author(s):  
Martin Potthast ◽  
Matthias Hagen ◽  
Benno Stein

No Web technology has undergone such an impressive evolution as Web search engines did and still do. Starting with the promise of "Bringing order to the Web" 1 by compiling information sources matching a query, retrieval technology has been evolving to a kind of "oracle machinery", being able to recommend a single source, and even to provide direct answers extracted from that source. Notwithstanding the remarkable progress made and the apparent user preferences for direct answers, this paradigm shift comes at a price which is higher than one might expect at first sight, affecting both users and search engine developers in their own way. We call this tradeoff "the dilemma of the direct answer"; it deserves an analysis which has to go beyond system-oriented aspects but scrutinize the way our society deals with both their information needs and means to information access. The paper in hand contributes to this analysis by putting the evolution of retrieval technology and the expectations at it in the context of information retrieval history. Moreover, we discuss the trade offs in information behavior and information system design that users and developers may face in the future.


Machines ◽  
2021 ◽  
Vol 9 (3) ◽  
pp. 66
Author(s):  
Tianci Chen ◽  
Rihong Zhang ◽  
Lixue Zhu ◽  
Shiang Zhang ◽  
Xiaomin Li

In an orchard environment with a complex background and changing light conditions, the banana stalk, fruit, branches, and leaves are very similar in color. The fast and accurate detection and segmentation of a banana stalk are crucial to realize the automatic picking using a banana picking robot. In this paper, a banana stalk segmentation method based on a lightweight multi-feature fusion deep neural network (MFN) is proposed. The proposed network is mainly composed of encoding and decoding networks, in which the sandglass bottleneck design is adopted to alleviate the information a loss in high dimension. In the decoding network, a different sized dilated convolution kernel is used for convolution operation to make the extracted banana stalk features denser. The proposed network is verified by experiments. In the experiments, the detection precision, segmentation accuracy, number of parameters, operation efficiency, and average execution time are used as evaluation metrics, and the proposed network is compared with Resnet_Segnet, Mobilenet_Segnet, and a few other networks. The experimental results show that compared to other networks, the number of network parameters of the proposed network is significantly reduced, the running frame rate is improved, and the average execution time is shortened.


Sensors ◽  
2021 ◽  
Vol 21 (6) ◽  
pp. 2117
Author(s):  
Hui Han ◽  
Zhiyuan Ren ◽  
Lin Li ◽  
Zhigang Zhu

Automatic modulation classification (AMC) is playing an increasingly important role in spectrum monitoring and cognitive radio. As communication and electronic technologies develop, the electromagnetic environment becomes increasingly complex. The high background noise level and large dynamic input have become the key problems for AMC. This paper proposes a feature fusion scheme based on deep learning, which attempts to fuse features from different domains of the input signal to obtain a more stable and efficient representation of the signal modulation types. We consider the complementarity among features that can be used to suppress the influence of the background noise interference and large dynamic range of the received (intercepted) signals. Specifically, the time-series signals are transformed into the frequency domain by Fast Fourier transform (FFT) and Welch power spectrum analysis, followed by the convolutional neural network (CNN) and stacked auto-encoder (SAE), respectively, for detailed and stable frequency-domain feature representations. Considering the complementary information in the time domain, the instantaneous amplitude (phase) statistics and higher-order cumulants (HOC) are extracted as the statistical features for fusion. Based on the fused features, a probabilistic neural network (PNN) is designed for automatic modulation classification. The simulation results demonstrate the superior performance of the proposed method. It is worth noting that the classification accuracy can reach 99.8% in the case when signal-to-noise ratio (SNR) is 0 dB.


2021 ◽  
Vol 11 (3) ◽  
pp. 1064
Author(s):  
Jenq-Haur Wang ◽  
Yen-Tsang Wu ◽  
Long Wang

In social networks, users can easily share information and express their opinions. Given the huge amount of data posted by many users, it is difficult to search for relevant information. In addition to individual posts, it would be useful if we can recommend groups of people with similar interests. Past studies on user preference learning focused on single-modal features such as review contents or demographic information of users. However, such information is usually not easy to obtain in most social media without explicit user feedback. In this paper, we propose a multimodal feature fusion approach to implicit user preference prediction which combines text and image features from user posts for recommending similar users in social media. First, we use the convolutional neural network (CNN) and TextCNN models to extract image and text features, respectively. Then, these features are combined using early and late fusion methods as a representation of user preferences. Lastly, a list of users with the most similar preferences are recommended. The experimental results on real-world Instagram data show that the best performance can be achieved when we apply late fusion of individual classification results for images and texts, with the best average top-k accuracy of 0.491. This validates the effectiveness of utilizing deep learning methods for fusing multimodal features to represent social user preferences. Further investigation is needed to verify the performance in different types of social media.


Sign in / Sign up

Export Citation Format

Share Document