A Graph-Based Approach for Mitigating Multi-Sided Exposure Bias in Recommender Systems

Fairness is a critical system-level objective in recommender systems that has been the subject of extensive recent research. A specific form of fairness is supplier exposure fairness, where the objective is to ensure equitable coverage of items across all suppliers in recommendations provided to users. This is especially important in multistakeholder recommendation scenarios where it may be important to optimize utilities not just for the end user but also for other stakeholders such as item sellers or producers who desire a fair representation of their items. This type of supplier fairness is sometimes accomplished by attempting to increase aggregate diversity to mitigate popularity bias and to improve the coverage of long-tail items in recommendations. In this article, we introduce FairMatch, a general graph-based algorithm that works as a post-processing approach after recommendation generation to improve exposure fairness for items and suppliers. The algorithm iteratively adds high-quality items that have low visibility or items from suppliers with low exposure to the users’ final recommendation lists. A comprehensive set of experiments on two datasets and comparison with state-of-the-art baselines show that FairMatch, although it significantly improves exposure fairness and aggregate diversity, maintains an acceptable level of relevance of the recommendations.

Research on labor consumption and rescuers’ physiological parameters progress in the case of practical training in a mobile training facility

An important factor in ensuring success in an intervention operation in dangerous environments lays in the training process, in conditions as close as possible to real intervention situations, as well as the manner of setting up rescue teams. In order to organize the intervention process as efficiently as possible, when setting up rescue teams we must take into account work consumption and physiological parameters achieved by members of the rescue teams during practical training carried out within the mobile training facility. In case of special incidents, the rescue intervention can take place in areas with small (limited) sections, spaces with high temperature, in low visibility conditions, which would make rescuers’ intervention even more difficult. Conditions mentioned above can be easily created within the rescuers’ mobile training facility, endowment of INSEMEX Petroșani, facility that through its equipment and facilities, allows monitoring of labour consumption and physiological parameters of rescuers throughout and in all training situations. The current paper presents training methods of intervention and rescue personnel within the mobile training facility and progress of work consumption / physiological parameters of rescuers throughout these practical exercises, with the aim of preparing intervention and rescue teams for situations close to real ones.

Research on rescuers’ training process in toxic / explosive / flammable environments

The paper presents the structuring of physical effort put in by rescuers during training, taking into account the activity specific psychosocial factors, structuring that aimed at streamlining the training process of intervention and rescue personnel in toxic / explosive / flammable environments. Training routes with various degrees of difficulty, allowing the simulation of intervention activities in horizontally and vertically confined spaces, low visibility, high temperature and humidity environments were analysed, for each the specific labour consumption being calculated. A dysfunctional phenomenon caused by the intervention and rescue activity is fatigue, perceived as a body reaction to readjust and restore its functions following intense or repeated body requests for energy consumption. Not only physical fatigue (which can be controlled through exercises) is specific for rescue and intervention activity but also mental fatigue. In addition to physical and technical training, psychosocial training of people who carry out intervention and rescue activities was followed in the training programs.

SURVEI SIDE SCAN SONAR DALAM PENELITIAN ARKEOLOGI BAWAH AIR DI PERAIRAN SUNGAI: STUDI KASUS PADA DAERAH ALIRAN SUNGAI BRANTAS [SIDE SCAN SONAR SURVEY IN RIVERINE UNDERWATER ARCHAEOLOGICAL RESEARCH: CASE STUDY IN THE BRANTAS BASIN]

Aplikasi metode geofisika menggunakan side scan sonar dalam penelitian arkeologi bawah air belum banyak dilakukan di Indonesia. Tulisan ini memaparkan penggunaan side scan sonar untuk pemetaan dasar sungai dan identifikasi tinggalan arkeologi di dasar sungai dalam penelitian “Sungai Brantas dalam Perspektif Lanskap Kultur Maritim”, serta interpretasi hasil survei side scan sonar tersebut dalam konteks kesejarahan. Selain itu, dalam tulisan ini akan dibahas potensi pengembangan penggunaan side scan sonar dalam penelitian arkeologi bawah air di Indonesia, terutama di perairan sungai. Akuisisi data dilakukan dengan menggunakan side scan sonar Starfish 450H dengan sistem posisi GNSS Trimble R8s. Sementara itu, interpretasi diperoleh dengan melakukan analisis terhadap data peta dan arsip Belanda untuk memahami konteks temporal dari objek yang dideteksi oleh alat side scan sonar. Survei berhasil menunjukkan sedimen di dasar sungai berupa lempung dan lanau, serta beberapa objek yang diduga sebagai bangkai kapal, yang diperkirakan berasal dari pasca abad ke-19 Masehi. Hasil survei side scan sonar menunjukkan tingkat akurasi cukup hingga tinggi dan dapat menjadi pendukung penelitian arkeologi bawah air yang efisien, terutama di perairan yang keruh. Side scan sonar survey as one of the geophysics methods is still scarcely applied in underwater archaeological research in Indonesia. This paper describes the application of side scan sonar survey in mapping riverbed and identifying underwater archaeological remains in the “Sungai Brantas in the Perspective of Maritime Cultural Landscape” project, as well as interpreting its historical context based on survey results. This paper also explores the development of utilizing side scan sonar in underwater archaeological research in Indonesia, particularly in rivers. Data was acquisitioned by using the side scan sonar Starfish 450H and GNSS Trimble R8s positioning system. The interpretation was drawn by analysing related Dutch old maps and archives to understand the historical context of the survey findings. The result shows clay and silt sediment covering most of the riverbed and a number of objects, possibly shipwrecks, estimated as from the nineteenth century. The survey result has a medium to high accuracy. Thus, this method is able to serve as an efficient instrument for underwater archaeological research, especially in the low-visibility waters.

Research on the usability of different machine learning methods in visibility forecasting

Haze pollution, mainly characterized by low visibility, is one of the main environmental problems currently faced by China. Accurate haze forecasts facilitate the implementation of preventive measures to control the emission of air pollutants and, thereby mitigate haze pollution. However, it is not easy to accurately predict the low visibility events induced by haze, which requires not only accurate prediction for weather elements, but also refined and real-time updated source emission inventory. In order to obtain reliable forecasting tools, this paper studies the usability of several popular machine learning methods, such as support vector machine, k-nearest neighbor, random forest, as well as several deep learning methods, on the visibility forecasting. Starting from the main factors related to visibility, the relationships between wind speed, wind direction, temperature, humidity, and visibility are discussed. Training and forecasting were performed using the machine learning methods. The accuracy of these methods in visibility forecasting was confirmed through several parameters (i.e., root-mean-square error, mean absolute error, and mean absolute percentage error). The results show that: (1) Among all meteorological parameters, wind speed was the best at reflecting the visibility change patterns; (2) RNN LSTM, and GRU methods performs almost equally well on short-term visibility forecasts(i.e. 1h, 3h, and 6h); (3) A classical machine learning method (i.e. the SVM) performs well in mid- and long-term visibility forecasts; (4) The machine learning methods also have a certain degree of forecast accuracy even for long time periods (e.g. of 72h).

Review on Parameterization Schemes of Visibility in Fog and Brief Discussion of Applications Performance

Low visibility, associated with fog, severely affects land, marine, and air transportation. Visibility is an important indicator to identify different intensities of fog; therefore, improving the ability to forecast visibility in fog is an urgent need for social and economic development. Establishing a proper visibility parameterization scheme is crucial to improving the accuracy of fog forecast operation. Considering various visibility impact factors, including RH, Nd, D, LWC, the parameterization formula of visibility in fog, as well as their performance in meteorology operation, are reviewed. Moreover, the estimated ability of the visibility parameterization formulas combined with the numerical model is briefly described, and their advantages and shortcomings are pointed out.

Climatology of blizzards over Schirmacher Oasis, east Antarctica

Weather in Antarctica is subject to frequent and sudden changes. Strong winds and blizzards dominate Antarctic weather. A combination of blowing snow, gale force wind and very low visibility is normally defined as blizzard. Meteorological data recorded at Indian Antarctic Station Maitri, in respect of blizzards recorded during the period 1990-2005 has been studied to find out climatological features of blizzards affecting Schirmacher Oasis. At Maitri the blizzard is mostly associated with extra-tropical storms and is normally preceded by precipitation. On average during the year about 21 blizzards affects the station for 45 days during the year. During the month of April to August 3 to 4 blizzards affects the station. Maximum number of blizzards occurs in the month of August with about 7 blizzard days. Average wind speed recorded during the blizzard is about 52 kt but it exceeded 100 kt on several occasions. The duration may vary from hours to days with average of 25 hours. Longest duration of 168 hours was recorded in June 1997. There are about 12 such occasions when blizzard lasted more than 72 hours. No correlation has been found between maximum wind speed and temperature rise during blizzard and the speed is also not correlated with pressure departure during the period.

Radiation fog over north India during winter from 1989-2004

Lkkj & bl 'kks/k-i= esa rsjg gokbZ vM~Mksa ds orZeku ekSle laca/kh vk¡dMksa dk mi;ksx djrs gq, mÙkjh Hkkjr esa fofdj.k ;qDr dksgjs dk v/;;u fd;k x;k gSA gky gh ds o"kksZa esa Hkkjr ds mÙkjh Hkkxksa esa dksgjs dh mRifÙk esa cgqr vf/kd o`f) ik;h xbZ gSA pw¡fd bafnjk xk¡/kh vUrjjk"Vªh; ¼vkbZ- th- vkbZ-½ gokbZ vM~Ms dh o"kZ 1997&1998] 1998&1999] 1999&2000] 2000&2001] 2001&2002] 2002&2003 vkSj 2003&2004 ds nkSjku dqy 900 mM+kuksa ds ekxZ cnys x,A blfy, bl gokbZ vM~Ms ds oSekfudh izpkyuksa ij iM+s dksgjs ds izfrdwy izHkko dk v/;;u fd;k x;k gSA bafnjk xk¡/kh gokbZ vM~Ms ij dSV&I, dSV&II vkSj dSV&III izpkyuksa ds fy, foekuksa dks mrkjus esa lgk;d vR;f/kd l{ke midj.k iz.kkyh vkbZ- ,y- ,l- miyC/k djkbZ xbZ gSA bl 'kks/k&i= esa xr iUnzg o"kksZa ds LdksiksxzkQksa ifjdfyr de n`’;rk okys izpkyuksa ds fy, vko’;d juos dh n`’; jsatksa vkj- oh- vkj- dh mi;ksfxrk ij fopkj&foe’kZ fd;k x;k gSA mixzg }kjk dksgjs ij fd, x, izs{k.kksa dk /kjkryh; izs{k.kksa ds lkFk lgh rkyesy ik;k x;k gSA mixzg ls izkIr gq, fp= bl ckr dk izek.k gSa fd o"kZ 1998&1999 ds nkSjku mŸkjh Hkkjr esa vR;f/kd l?ku dksgjk vjc lkxj esa cus izpaM pØokr ls vR;f/kd ek=k esa vknzZrk ds izokg ds dkj.k cuk FkkA bl 'kks/k-i= esa bafnjk xk¡/kh gokbZ vM~Mk] y[kuÅ gokbZ vM~Mk] okjk.klh gokbZ vM~Mk vkSj ve`rlj gokbZ vM~Mk ij dksgjs ds nkSjku vf/kdre rkieku vkSj lkisf{kd vknzZrk dh folaxfr;ksa ds e/; laca/k dk irk yxkus dk Hkh iz;kl fd;k x;k gSA Radiation fog over north India has been studied using current weather data of thirteen airports. There has been a tremendous increase in the fog formation over northern parts of India in recent years. An attempt has been made to study the adverse impact of fog on aeronautical operations at Indira Gandhi International (I.G.I.) airport as total number of flights diverted during 1997-98, 1998-99, 1999-2000, 2000-01, 2001-02, 2002-03 and 2003-04 were about 900. I.G.I. airport is provided with a very efficient Instrument Landing System (ILS) for Cat-I, Cat-II and Cat-III operations. The utility of Runway Visual Ranges (RVRs) required for low visibility operations, calculated from skopographs, for the last fifteen years, has been discussed. Satellite observations on fog have been found to be in fair agreement with the surface observations. Most catastrophic fog formations, which occurred over north India during 1998-99, were found to be due to the enormous amount moisture flow from a severe cyclone formed in the Arabian Sea as evidenced in satellite imagery. An attempt has also been made to establish a relation of maximum temperature and Relative Humidity anomaly with the duration of fog at I.G.I. airport, Lucknow airport, Varanasi airport and Amritsar airport.

Validating a model of architectural hazard visibility with low-vision observers

Pedestrians with low vision are at risk of injury when hazards, such as steps and posts, have low visibility. This study aims at validating the software implementation of a computational model that estimates hazard visibility. The model takes as input a photorealistic 3D rendering of an architectural space, and the acuity and contrast sensitivity of a low-vision observer, and outputs estimates of the visibility of hazards in the space. Our experiments explored whether the model could predict the likelihood of observers correctly identifying hazards. In Experiment 1, we tested fourteen normally sighted subjects with blur goggles that simulated moderate or severe acuity reduction. In Experiment 2, we tested ten low-vision subjects with moderate to severe acuity reduction. Subjects viewed computer-generated images of a walkway containing five possible targets ahead—big step-up, big step-down, small step-up, small step-down, or a flat continuation. Each subject saw these stimuli with variations of lighting and viewpoint in 250 trials and indicated which of the five targets was present. The model generated a score on each trial that estimated the visibility of the target. If the model is valid, the scores should be predictive of how accurately the subjects identified the targets. We used logistic regression to examine the correlation between the scores and the participants’ responses. For twelve of the fourteen normally sighted subjects with artificial acuity reduction and all ten low-vision subjects, there was a significant relationship between the scores and the participant’s probability of correct identification. These experiments provide evidence for the validity of a computational model that predicts the visibility of architectural hazards. It lays the foundation for future validation of this hazard evaluation tool, which may be useful for architects to assess the visibility of hazards in their designs, thereby enhancing the accessibility of spaces for people with low vision.

Terrain visibility impact on the preparation of landslide inventories: some practical cases

Landslide inventories are used for multiple purposes including landscape characterisation and monitoring, and landslide susceptibility, hazard and risk evaluation. Their quality can depend on the data and the methods with which they were produced. In this work we evaluate the effects of a variable visibility of the territory to map on the spatial distribution of the information collected by four landslide inventories prepared using different approaches in two study areas. The method first classifies the territory in areas with different visibility levels from the paths (roads) used to map landslides, and then estimates the landslide density reported in the inventories into the different visibility classes. Our results show that 1) the density of the information is strongly related to the visibility in inventories obtained through fieldwork, technical reports and/or newspapers, where landslides are under-sampled in low visibility classes; and 2) the inventories obtained by photo-interpretation of images suffer from a marked under representation of small landslides close to roads or infrastructures. We maintain that the proposed procedure can be useful to evaluate the quality of landslide inventories and then properly orient their use.