Historical Behavior Research Articles

Extreme indices of temperature and precipitation in South America: trends and intercomparison of regional climate models

Regional Climate Models (RCMs) provide climate information required for evaluating vulnerability, impacts, and adaptation at finer scales than their global driving models. As they explicitly resolve the basic conservation and state equations, they solve physics with more detail, conserving teleconnection of larger scales provided by Global Climate Models (GCMs). In South America (SA), the regional simulations have been historically evaluated principally on climatological aspects, but the representativeness of extremes still needs a more profound assessment. This study aims to analyze three RCMs (RegCM4-7, REMO2015, and Eta) driven by different GCMs in SA, focusing on their capacity to reproduce extreme historical indices of daily precipitation and temperature. The indices of maximum consecutive 5 days precipitation (Rx5day), Consecutive Dry Days (CDD), daily maximum and minimum annual temperature (TXx and TNn, respectively) were evaluated regarding the historical spatio-temporal variability and trends. Furthermore, their projections for the 2071–2099 period, under the Representative Concentration Pathway 8.5 scenario, were analyzed. The historical behavior of RCMs (1981–2005) was compared with two gridded products: Climate Prediction Center (CPC) and agrometeorological indicators derived from the fifth generation of global reanalysis produced by the ECMWF (AgERA5), previously compared with records from meteorological stations to evaluate them. The results show that the highest differences within the gridded products and stations were observed in the regions with more scarce surface stations (North and West of SA) and with complex topography (The Andes Cordillera), being more pronounced in the precipitation-based indices. We found that RCMs generally show more agreement in the spatial variability than in the inter-annual variability for all the indices and SA regions. When analyzing the observed trends, all models better reproduced the long-term variability of extreme temperature indices than those of rainfall. More disagreement was observed for Rx5day and CDD indices trends, including substantial spatial heterogeneities in both magnitude and sign of tendency. Climate change projections exhibited significant agreement to warmer conditions in TXx and TNn, but precipitation signals differed between RCMs and the driving GCM within each regional model. Maximum dry spells are expected to increase in almost all SA regions, whereas the climate change signals in extreme precipitation events are more consistent over southeastern SA (northern and southwestern SA), with positive (negative) changes by the end of the century.

Exploiting User Preferences for Multiscenarios in Query-Less Search

Online travel platforms (OTPs), for example, booking.com, Ctrip.com, and Fliggy, deliver travel experiences to online users by providing travel-related products. Hotel recommendation is significantly important for OTPs since hotel bookings account for almost half of the travel expenses and hotel products generate more than half of OTP’s revenues. More than 58% Fliggy users may choose to use query-less hotel searches to find candidate hotels, where no additional keywords are given except the expected check-in date and travel destination city. Thus, how to recommend hotels to traveler users is important and challenging. In this article, we explore the unique characteristics of query-less hotel users and propose a novel multiscenario query-less search network (MSQS). According to their searching date, expected check-in date, current city, and expected hotel city, MSQS groups users’ behaviors (e.g., click, purchase, search) into four scenario groups, namely today-local, today-nonlocal, future-local, and future-nonlocal. The key components of MSQS are the global expert, the scenario expert, and the feedback expert. The global expert learns common features among different scenarios and extracts the feature interactions between context, users, and hotels. The scenario expert utilizes multilayer perception to learn the differentiating features between scenarios. The feedback expert learns users’ preferences for hotels in different scenarios through their historical behaviors, and a scenario interest extractor is carefully designed to enhance attention across scenarios and behaviors. An offline experiment on the Fliggy production dataset with over 8 million users and 0.49 million travel items and an online A/B test both show that MSQS effectively predicts users’ hotel booking intentions.

A Flight fare Prediction Using Machine Learning

Abstract: While airlines (the sellers) always work to increase their revenue by changing pricing for the same service, air travellers (the buyers) frequently search for the ideal time of year to buy flights in order to save as much money as possible. The choice to raise or lower tickets at various points leading up to departure dates can be made by the sellers based on all the relevant data, such as historical sales, market demand, consumer profile, and behaviour. The buyers, on the other hand, have limited access to data to help them decide whether to delay or make a quick flight purchase. In this study, we suggest a new model that might assist the purchaser in anticipating price movements even in the absence of official airlines. Our results showed that the suggested model, despite lacking several essential components, such as the number of unsold seats on flights, can forecast trends as well as actual changes in airfare up to the departure dates using public airfare data that is readily available online. We also determined the characteristics that have the biggest effects on changes in airfare

A Flight Fare Prediction Using Machine Learning

Abstract: While airlines (the sellers) always work to increase their revenue by changing pricing for the same service, air travellers (the buyers) frequently search for the ideal time of year to buy flights in order to save as much money as possible. The choice to raise or lower tickets at various points leading up to departure dates can be made by the sellers based on all the relevant data, such as historical sales, market demand, consumer profile, and behaviour. The buyers, on the other hand, have limited access to data to help them decide whether to delay or make a quick flight purchase. In this study, we suggest a new model that might assist the purchaser in anticipating price movements even in the absence of official airlines. Our results showed that the suggested model, despite lacking several essential components, such as the number of unsold seats on flights, can forecast trends as well as actual changes in airfare up to the departure dates using public airfare data that is readily available online. We also determined the characteristics that have the biggest effects on changes in airfare.

Are people less generous after a family member gives to charity? The interaction of self-construal and relationship type

Extant research shows that people become more generous after learning about the charitable giving of in-group members (e.g., family or friends). The present research demonstrates an exception: While a friend’s donation increases both the interdependents’ and the independents’ charitable giving, a family member’s donation decreases the interdependents’ but increases the independents’ charitable giving. Six studies, including a field experiment and an online survey that captures historical behavior, provide converging evidence in support of an interaction effect of consumers’ self-construal and their relationship type with previous donors on their generosity, showing that the interdependents become less generous after a family member (vs friend) donates. Furthermore, we demonstrate that the interaction effect is driven by the perception of shared responsibility and that this interaction effect is weakened when consumers’ personal responsibility is enhanced or when the donation currency is time rather than money. Finally, the theoretical and practical implications of the work are discussed.

A Large Group Decision Making Method Considering Experts’ Non-cooperative Behavior for Investment Selection of Renewable Energy Projects

The rapid expansion of renewable energy has attracted the attention of investors, which makes the evaluation of renewable energy projects a momentous issue. As the investment selection of renewable energy projects requires the joint discussion of experts from different professional backgrounds (such as energy, transportation, construction, economy, environment, etc.), it belongs to the category of large group decision-making (LGDM). Therefore, this paper is devoted to propose a novel LGDM method considering experts’ non-cooperative behavior for investment selection of renewable energy projects. First, considering that the complexity of renewable energy projects makes it difficult for experts to express their views in a single linguistic word, the hesitant fuzzy linguistic term set is used as the tool for expert evaluation in this paper. Second, since the assessment information provided by experts from different fields are often heterogeneous, a consensus-reaching process with a feedback mechanism is introduced which comprehensively considers three reliable sources: the experts’ trust relationship in the social trust network, the consensus contribution in the subgroup and the opinions’ similarity among experts. Further, to improve the efficiency and rationality of decision-making, an experts’ historical adjustment data-based non-cooperative behavior management method is proposed. Finally, the effectiveness and innovation of the proposed method are verified by a case of renewable energy power generation project investment selection in Qingdao, China and a series of comparative analysis.

DACSR: Decoupled-Aggregated End-to-End Calibrated Sequential Recommendation

Sequential recommendations have made great strides in accurately predicting the future behavior of users. However, seeking accuracy alone may bring side effects such as unfair and overspecialized recommendation results. In this work, we focus on the calibrated recommendations for sequential recommendation, which is connected to both fairness and diversity. On the one hand, it aims to provide fairer recommendations whose preference distributions are consistent with users’ historical behaviors. On the other hand, it can improve the diversity of recommendations to a certain degree. But existing methods for calibration have mainly relied on the post-processing on the candidate lists, which require more computation time in generating recommendations. In addition, they fail to establish the relationship between accuracy and calibration, leading to the limitation of accuracy. To handle these problems, we propose an end-to-end framework to provide both accurate and calibrated recommendations for sequential recommendation. We design an objective function to calibrate the interests between recommendation lists and historical behaviors. We also provide distribution modification approaches to improve the diversity and mitigate the effect of imbalanced interests. In addition, we design a decoupled-aggregated model to improve the recommendation. The framework assigns two objectives to two individual sequence encoders, and aggregates the outputs by extracting useful information. Experiments on benchmark datasets validate the effectiveness of our proposed model.

Anticipated versus Actual Effects of Platform Design Change: A Case Study of Twitter's Character Limit

The design of online platforms is both critically important and challenging, as any changes may lead to unintended consequences, and it can be hard to predict how users will react. Here we conduct a case study of a particularly important real-world platform design change: Twitter's decision to double the character limit from 140 to 280 characters to soothe users' need to "cram" or "squeeze" their tweets, informed by modeling of historical user behavior. In our analysis, we contrast Twitter's anticipated pre-intervention predictions about user behavior with actual post-intervention user behavior: Did the platform design change lead to the intended user behavior shifts, or did a gap between anticipated and actual behavior emerge? Did different user groups react differently? We find that even though users do not "cram" as much under 280 characters as they used to under 140 characters, emergent "cramming" at the new limit seems to not have been taken into account when designing the platform change. Furthermore, investigating textual features, we find that, although post-intervention "crammed" tweets are longer, their syntactic and semantic characteristics remain similar and indicative of "squeezing". Applying the same approach as Twitter policy-makers, we create updated counterfactual estimates and find that the character limit would need to be increased further to reduce cramming that re-emerged at the new limit. We contribute to the rich literature studying online user behavior with an empirical study that reveals a dynamic interaction between platform design and user behavior, with immediate policy and practical implications for the design of socio-technical systems.

Self-Attention Mechanism Enhanced User Interests Modeling for Personalized Recommendation Services in Cyber-Physical-Social Systems

Cyber-Physical-Social Systems (CPSS) provide great value to our lives, but they also cause data overload problems. Data-driven personalized recommendation service is one of the most efficient means to solve such problems, which is currently receiving wide attention from research and industrial communities. The most important task of personalized recommender systems is to predict the click-through rate of given items, which is especially true for personalized advertisement recommendation systems. Recently, a number of deep click-through models have been proposed, which obtain low-dimensional dense embedding vectors of features, and then concatenate together and input into multi-layer perceptron to learn the nonlinear relationship between the features. However, the existing models don't dig deep enough into the user preferences and habits in users’ behavior history. In this paper, we propose a new model: Self-attention based Deep Neural Network (DeepSA), which addresses this issue by constructing Ad-related graph and training graph embedding vectors to enhance the representation of the advertisement for capturing user interests, and learns the internal correlation between user behaviors via the self-attention mechanism, which better explores interests and preferences hidden in users’ historical behaviors. Experiments on two public datasets and an industrial dataset demonstrate the proposed method outperforms the state-of-the-art models.

Some examples for stable and historic behavior in replicator equations

The evolutionary dynamics of zero-sum and non zero-sum games under replicator equations could be drastically different from each other. In zero-sum games, heteroclinic cycles naturally occur whenever the species of the population supersede each other in cyclic fashion (like for the Rock–Paper–Scissors game). In this case, the highly erratic oscillations may cause the divergence of the time averages. In contrast, it is a common belief that all “reasonable” replicator equations of non-zero sum games satisfy “The Folk Theorem of Evolutionary Game Theory” which asserts that (i)a Nash equilibrium is a rest point; (ii)a stable rest point is a Nash equilibrium; (iii)a strictly Nash equilibrium is asymptotically stable; (iv)any interior convergent orbit evolves to a Nash equilibrium. In this paper, we propose two distinct classes of replicator equations generated by Schur-convex potential functions which exhibit two opposing phenomena: stable/predictable and historic/unpredictable behavior. In the latter case, the time averages of the orbit will slowly oscillate during the evolution of the system and do not converge to any limit. This will eventually cause the divergence of higher-order repeated time averages.

Analysis of Historical Patenting Behavior and Patent Characteristics of Computer Architecture Companies—Part V: References

In prior parts of this series, I analyzed the 1) number of all issued patents and issued computer architecture patents; 2) the prosecution time and effective patent term; 3) the number of claims, breakdown of independent and dependent claims, and effect that excess claim fees had on the numbers of total and independent claims; and 4) the type of claims (apparatus, method, or Beauregard), and effect that the Supreme Court’s decision in Alice v. CLS Bank had on the number of independent and dependent method claims, for patents that were issued to 18 leading computer architecture companies and filed between 1996 and 2020. This article examines the 1) number of “backward” citations to previously issued/published U.S. patents and U.S. patent publications, foreign patents, and Other References and 2) the number of “forward” citations to a company’s patent by another U.S. patent or U.S. patent publication.

Reputation-Based Blockchain for Spatial Crowdsourcing in Vehicular Networks

The sharing of high-quality traffic information plays a crucial role in enhancing the driving experience and safety performance for vehicular networks, especially in the development of electric vehicles (EVs). The crowdsourcing-based real-time navigation of charging piles is characterized by low delay and high accuracy. However, due to the lack of an effective incentive mechanism and the resource-consuming bottleneck of sharing real-time road conditions, methods to recruit or motivate more EVs to provide high-quality information gathering has attracted considerable interest. In this paper, we first introduce a blockchain platform, where EVs act as the blockchain nodes, and a reputation-based incentive mechanism for vehicular networks. The reputations of blockchain nodes are calculated according to their historical behavior and interactions. Further, we design and implement algorithms for updating honest-behavior-based reputation as well as for screening low-reputation miners, to optimize the profits of miners and address spatial crowdsourcing tasks for sharing information on road conditions. The experimental results show that the proposed reputation-based incentive method can improve the reputation and profits of vehicle users and ensure data timeliness and reliability.

RecNet: A Resource-Constraint Aware Neural Network for Used Car Recommendation

Resource constraints, e.g., limited product inventory or financial strength, may affect consumers’ choices or preferences in some recommendation tasks but are usually ignored in previous recommendation methods. In this paper, we aim to mine the cue of user preferences or intention in the case of resource-constraint recommendation tasks. For this purpose, we specifically build a largely used car transaction dataset possessing resource-constraint characteristics. Accordingly, we propose a resource-constraint-aware network to predict the user’s future interaction based on dynamic connections between users and items. To describe the user-item connection dynamically, mutually recursive recurrent neural networks (MRRNNs) are introduced to capture long-term interactive dependencies, and effective representations of users and items are obtained. To further consider the resource constraint, a resource-constraint branch is built to explore resource variation’s influence on user preferences. Finally, mutual information is introduced to measure the similarity between the future user action and fused historical behavior features to predict future interaction. The fused features come from both MRRNNs and resource-constraint branches. We test the performance on the built used car transaction dataset and the Tmall dataset, and the experimental results verify the effectiveness of our framework.

Joint Attention Networks with Inherent and Contextual Preference-Awareness for Successive POI Recommendation

Nowadays recording and sharing personal lives using mobile devices on the Internet is becoming increasingly popular, and successive POI recommendation is gaining growing attention from academia and industry. In mobile scenarios, multiple influencing factors including the diversity of user preferences, the changeability of user behavior and the dynamic of spatiotemporal context bring great challenges to the POI recommender system. In order to accurately capture both the stable and the contextual preferences of mobile users in dynamic contexts, we propose a fusion framework JANICP (Joint Attention Networks with Inherent and Contextual Preferences) for successive POI recommendation by jointly training an offline/nearline user inherent interest perception model and an online user contextual interest prediction model. The offline model is trained based on the global historical behavior data to achieve stable interest representation, while the online model is trained based on the instantly selected context-sensitive data to achieve dynamic interest perception. An attention aggregation and matching module is used to fully connect the two kinds of preference representations and generate the final POI recommendation. Extensive experiments were conducted on three real datasets and experimental results show that the proposed JANICP outperforms existing state-of-the-art methods.

SCF: Structured collaborative filtering with heterogeneous implicit feedback

Recommendation systems aim to analyze users’ historical behaviors to recommend items that suit their preferences. In the real world, users’ feedback is usually heterogeneous, such as purchases and examinations, and there are two fundamental challenges, namely the heterogeneity of users’ implicit feedback and the uncertainty of users’ preferences. Previous works have attempted to exploit heterogeneous implicit feedback, but most of them are based on traditional preference assumptions and may not fully learn users’ preferences. To overcome this limitation, we propose a novel solution, i.e., structured collaborative filtering (SCF), which can effectively utilize the complementarity of heterogeneous feedback. Specifically, we address the heterogeneity challenge by designing two types of structured information, i.e., a set of examined items and a group of neighboring users, which integrate examinations into the task of learning users’ purchase preferences. Furthermore, we reduce the uncertainty of users’ preferences by leveraging two preference assumptions defined on randomly constructed item-sets or user-groups. On this basis, we derive four algorithms, including SCF(i,i), SCF(u,i), SCF(i,u), and SCF(u,u), which take advantage of pointwise preference learning to accurately model the latent representations of users and items. Extensive experimentation on three real-world datasets demonstrates that our SCF outperforms the state-of-the-art methods.

Personalized Dynamic Attention Multi-task Learning model for document retrieval and query generation

In the actual retrieval scenario, queries from different users tend to have different retrieval intentions. Accurately understanding users’ retrieval intentions is a fundamental challenge for search engines. In this paper, we propose a Personalized Dynamic Attention Multi-task Learning model (PDAML) to solve this problem, which can clarify the user’s retrieval intention and generate a personalized document list. Specifically, we design a personalized retrieval model that can learn user retrieval preferences based on the user’s historical behavior. In addition, we propose an ad-hoc model based on multi-task learning to train query generation tasks and document retrieval tasks to enhance query representation and integrate document-aware interactive information into query representation through a dynamic attention mechanism. We can generate query representations with intent information and rank candidate documents through the interaction of the two sub-models. Experimental results on the publicly available AOL and SogouQ datasets demonstrate the effectiveness of PDAML, which improves the retrieval accuracy and personalized experience for users.

A Location Recall Strategy for Improving Efficiency of User-Generated Short Text Geolocalization

Currently, users’ daily lives are closely related to social sites. Geolocalizing user-generated short texts (UGSTs) can provide services for the wider fields of business and society. The existing works focus on improving the accuracy of geolocalization algorithms and reducing the prediction error. However, little work has been done on improving the computing efficiency of geolocalization algorithms. When dealing with large-scale data, the existing algorithm efficiency is greatly reduced. To solve this problem, we summarize three modes of user access to point of interests (PoIs), further propose a variety of location recall methods, and then introduce these geolocation recall methods into the existing algorithm. Geolocation recall generates a candidate location set for each user according to the user’s historical behavior records to reduce the search space of the geolocalization algorithm and improve its efficiency. We carried out experiments on the ground-truth datasets, compared the performance of various location recall methods, and verified the excellent location recall performance of three geolocalization algorithms. The experimental results show that location recall shortens the calculation time of the three algorithms by approximately 70% and further improves the accuracy of these algorithms.

Analysis of Social Media Algorithm Recommendation System

Once, the Internet was considered an important force to promote social public welfare activities and spread social justice. Today, the driving force of the Internet on the spread of negative social information and polarisation also exists. Social media platforms use this kind of method to customise personalised content distribution schemes by mining and analysing individual historical behaviour data based on algorithms. This behaviour of the platform brings about efficient and accurate information matching. At the same time, the content recommended by the algorithm is also controversial due to the risk of privacy violations and other issues. This article will analyse the advantages and disadvantages of social media platform algorithms and recommendation systems for citizens.

Music Recommendation Based on “User-Points-Music” Cascade Model and Time Attenuation Analysis

Music has an increasing impact on people’s daily lives, and a sterling music recommendation algorithm can help users find their habitual music accurately. Recent research on music recommendation directly recommends the same type of music according to the specific music in the user’s historical favorite list. However, users’ behavior towards a certain cannot reflect the preference for this type of music and possibly provides music the listener dislikes. A recommendation model, MCTA, based on “User-Point-Music” structure is proposed. By clustering users’ historical behavior, different interest points are obtained to further recommend high-quality music under interest points. Furthermore, users’ interest points will decay over time. Combined with the number of music corresponding to each interest point and the liking degree of each music, a multi-interest point attenuation model is constructed. Based on the real data after desensitization and encoding, including 100,000 users and 12,028 pieces of music, a series of experimental results show that the effect of the proposed MCTA model has improved by seven percentage points in terms of accuracy compared with existing works. It came to the conclusion that the multi-interest point attenuation model can more accurately simulate the actual music consumption behavior of users and recommend music better.

Multisource Information Fusion Algorithm for Personalized Tourism Destination Recommendation

In this paper, the existing scenic spot recommendation algorithms ignore the implicit trust and trust transmission of users when dealing with user relationships, and the lack of historical browsing behavior data of users in new city scenes leads to an inaccurate recommendation. In this paper, a personalized scenic spot recommendation method combining user trust relationship and tag preference is proposed. Firstly, the trust degree is introduced when the recommendation quality is poor only considering the similarity of users. By mining the implicit trust relationship of users, the problem that the existing research cannot make recommendations when the direct trust is difficult to obtain is solved, and the data sparsity and cold start problems are effectively alleviated. Secondly, in the process of user interest analysis, the relationship between scenic spots and tags is extended to the relationship among users, scenic spots and tags, and users’ interest preferences are decomposed into long-term preferences for different scenic spots tags, which effectively alleviates the problem of poor recommendation quality when users’ historical tour records are lacking. The personalized tourism recommendation method proposed in this paper effectively integrates many features of social networks and effectively alleviates the problems of data sparseness and feature learning in tourism recommendation based on social networks by using vectorization and deep learning technology. Its research has very important usage scenarios and commercial value in the tourism industry. This model can efficiently mine the association rules between scenic spots in multisource information data. The experimental results show that mining the correlation between the scenic spots selected by tourists can provide effective information for tourism decision-making.