A Hybrid Method Based on HITS for Literature Recommendation

2011 ◽  
Vol 55-57 ◽  
pp. 1636-1641 ◽  
Author(s):  
Pei Ying Zhang ◽  
Ya Jun Du ◽  
Chang Wang

In this paper we propose a hybrid method of literature recommendation in the academic community. First, we refer the objective recommendation based on HITS algorithm by constructing a directed graph according to the literature citation relation and then select the articles considering the authority and hub score of each article synthetically and add them to the recommendation list. This can narrow the recommendation scope and give a more authoritive recommendation. Second, the subjective recommendation is based on collaborative filtering by comparing the ratings of other similar users for the objects in recommendation list. The difference is we discover the similar user by clustering them. And the experiment shows the method can provide better recommendation results and is timesaving.

2019 ◽  
Vol 28 (12) ◽  
pp. 1950076
Author(s):  
Thomas Fleming ◽  
Joel Foisy

A directed graph [Formula: see text] is intrinsically linked if every embedding of that graph contains a nonsplit link [Formula: see text], where each component of [Formula: see text] is a consistently oriented cycle in [Formula: see text]. A tournament is a directed graph where each pair of vertices is connected by exactly one directed edge. We consider intrinsic linking and knotting in tournaments, and study the minimum number of vertices required for a tournament to have various intrinsic linking or knotting properties. We produce the following bounds: intrinsically linked ([Formula: see text]), intrinsically knotted ([Formula: see text]), intrinsically 3-linked ([Formula: see text]), intrinsically 4-linked ([Formula: see text]), intrinsically 5-linked ([Formula: see text]), intrinsically [Formula: see text]-linked ([Formula: see text]), intrinsically linked with knotted components ([Formula: see text]), and the disjoint linking property ([Formula: see text]). We also introduce the consistency gap, which measures the difference in the order of a graph required for intrinsic [Formula: see text]-linking in tournaments versus undirected graphs. We conjecture the consistency gap to be nondecreasing in [Formula: see text], and provide an upper bound at each [Formula: see text].


2021 ◽  
Vol 14 (1) ◽  
pp. 821-840
Author(s):  
Adam Radzimski ◽  
Jedrzej Gadzinski

Investments in light rail transit (LRT) have become increasingly popular solutions to promote sustainable urban transportation. However, their impacts on cities are still subject to discussion in the academic community. There is a clear need to better understand the potential impacts of LRT projects, particularly in contexts other than major cities. In this study, we focus on the Olsztyn tram project, which has been implemented in a city of 173,000 residents situated in northeastern Poland. The paper combines different perspectives and data sources, including a study of residents’ stated preferences concerning travel behavior and modelling of housing price effects using the difference-in-differences approach. Our results suggest that the Olsztyn tram project led to a moderate change in travel behavior by increasing the frequency of public transport use but did not result in a substantial shift away from car commuting. Concerning the property market, a decline in prices was observed during the construction phase, but no statistically significant effects were found after completion.


Author(s):  
Paul Lauter

When you feel yourself beginning to slide down a cliff, you are not likely to think too hard about what it is you grab to stop the fall. But the choice of handholds makes a difference—the difference between continuing to plunge and holding on long enough to plant your feet. As you descend, what seems a vine turns out to be a viper, and what seems a solid trunk proves rootless and tears away. So it is as faculty have contended with the growing shelf of studies criticizing, occasionally analyzing, and mostly prescribing for, higher education. We feel the structure, the norms of our profession, shifting and sliding beneath our feet. We reach for a handhold, a point of stability, and discover, alas, that there’s little that is reliable, much that is frail and fragile. Three of the mid-1980s higher education studies1 were among the opening shots in what has become an extended battle over the character and quality of the institutions in which professors work, as well as over what exactly it is that faculty and staff do. One could, of course, dismiss these and more recent studies, perhaps citing their manifold banalities as sufficient reason for indifference. Or, as faculty, we could acquiesce, agreeing to such changes as the reformists are able to compel, but doing little more than what is necessary to protect our turf. Either course is rationally defensible. Neither is advisable for the academic community. It seems to me that either indifference or generalized resistance would be mistaken—for at least two reasons. First, this has proven to be an unusually strong tide of reform, and even now, half a decade later, it seems still to be waxing. Even from the perspective of strict self-interest, not an unfamiliar ground for academics to stand upon, it would be dangerous to ignore what is a continuing effort to reshape the character of our work and lives. Second, the drive to reform college education presents faculty and staff with an opportunity to shape the direction of change, and in particular to raise what none of these reports really contends with: What political values, what economic forms, what social objectives do we really wish to pursue?


2013 ◽  
Vol 677 ◽  
pp. 412-417
Author(s):  
Suppatoomsin Chompoo ◽  
Srikaew Arthit

This paper presents a hybrid method for vehicle detection from CCTV captured image. In order to overwhelm such complex details of the color image, the system combines artificial intelligence techniques to achieve automatic vehicle detection. These are techniques 2D principal component analysis (2DPCA), Fuzzy adaptive resonance theory (Fuzzy ART), genetic algorithm (GA) and self-organizing map. The proposed system can detect different vehicle sizes from different proportional image area. Bilinear interpolation is used to resize each proportional image area to vehicle feature matrix. The proposed system can detect various types of vehicles from the difference image background.


2021 ◽  
Author(s):  
Jason Blazakis ◽  
Colin Clarke

The global far right is extremely broad in nature and far from monolithic. While the “far right” is often used as an umbrella term, using the term runs the risk of over-simplifying the differences and linkages between white supremacist, anti-immigration, nativist, and other motivating ideologies. These beliefs and political platforms fall within the far-right rubric, and too often the phrase presents a more unified image of the phenomena than is really the case. In truth, the “far right” and the individual movements that comprise it are fragmented, consisting of a number of groups that lack established leadership and cohesion. Indeed, these movements include chauvinist religious organizations, neo-fascist street gangs, and paramilitary organs of established political parties. Although such movements largely lack the mass appeal of the interwar European radical right-wing extreme, they nevertheless can inspire both premeditated and spontaneous acts of violence against perceived enemies. This report is intended to provide policymakers, practitioners, and the academic community with a roadmap of ongoing shifts in the organizational structures and ideological currents of radical right-wing extremist movements, detailing the difference between distinct, yet often connected and interlaced echelons of the far right. In particular, the report identifies and analyzes various aspects of the broader far right and the assorted grievances it leverages to recruit, which is critical to gaining a more nuanced understanding of the potential future trajectory of these movements.


2021 ◽  
Vol 2021 ◽  
pp. 1-14
Author(s):  
Muhammad Shaheen

The definition of research appears to be a controversial subject in the academic community. It is almost becoming apparent that research remains a key element in master’s and doctoral degrees. Originality is, to varying degrees, the primary concern for the genuineness of research, but there is a difference of opinion on the concept of originality. In practice, a subjective analysis of originality is performed at the time of the assessment of the said degrees’ theses; therefore, the resulting evaluation is affected by the difference of opinion on the concept of originality. It has also emerged that the concept of originality in research is also marginally unique in various areas. This study is focused upon reaching a point of agreement for the definition of originality in the theses/dissertation of masters and doctorate degrees in engineering only. The outcome of the research should be something new and originally contribute to the body of knowledge. This study focused on to arrive at an accurate definition of originality in university degrees in the field of engineering. An online survey was designed and carried out on the basis of evidence and expert opinion. The survey was distributed amongst engineering peers. The results of the survey are systematically summed up in the study.


Author(s):  
Hanfei Zhang ◽  
Yumei Jian ◽  
Ping Zhou

: A class correlation distance collaborative filtering recommendation algorithm is proposed to solve the problems of category judgment and distance metric in the traditional collaborative filtering recommendation algorithm, which is using the advantage of the distance between the same samples and the class related distance. First, the class correlation distance between the training samples is calculated and stored. Second, the K nearest neighbor samples are selected, the class correlation distance of training samples and the difference ratio between the test samples and training samples are calculated respectively. Finally, according to the difference ratio, we classify the different types of samples. The experimental result shows that the algorithm combined with user rating preference can get lower MAE value, and the recommendation effect is better. With the change of K value, CCDKNN algorithm is obviously better than KNN algorithm and DWKNN algorithm, and the accuracy performance is more stable. The algorithm improves the accuracy of similarity and predictability, which has better performance than the traditional algorithm.


Author(s):  
Aras Satria Agusta

Introduction. This article is entitled "Analysis of webometrics content in the Syiah Kuala University repository and the University of North Sumatra". The purpose in writing is to look at the quality of institutional repositories based on webometrics content, which has an impact on the ranking of webometrics 2020. Data Collection Method. In this article the authors use a descriptive quantitative approach, while observations are made by observing and analyzing search results on webometrics indicator devices systematically and in a standardized manner. existing indicators, data generated for each webometrics content indicator against size indicators, visibility indicators, rich file indicators, scholar indicators on institutional repository websites through search engines and normalized. Result and Discussions. The results of this study are that each indicator of size, visibility, rich file and scholar in the repository of the University of North Sumatra is superior to Syiah Kuala University with the total ranking of the December 2019 webometrics indicator is 3.56506 while Syiah Kuala University with a total value of 0.83811 . Then the difference in the total rating in the repository is 2.72695, which allows a change in the ranking of the two universities. Conclusions. From the results of the repository ranking, the University of North Sumatra was superior with a score of 3.56506 while Syiah Kuala University with a total score of 0.83811. Then the difference in the total ranking value in the repository is 2,72695 which has an impact on each achievement of activeness with loyal members of the community in developing institutional repositories. From this, the academic community of each campus should encourage their scientific works to be published on the repository website they already have, while students submit their scientific work in the form of a paper or final project to the library and then processed and disseminated on the repository website.


2019 ◽  
Vol 06 (03) ◽  
pp. 363-376 ◽  
Author(s):  
Gharbi Alshammari ◽  
Stelios Kapetanakis ◽  
Abdullah Alshammari ◽  
Nikolaos Polatidis ◽  
Miltos Petridis

Recommender systems help users find relevant items efficiently based on their interests and historical interactions with other users. They are beneficial to businesses by promoting the sale of products and to user by reducing the search burden. Recommender systems can be developed by employing different approaches, including collaborative filtering (CF), demographic filtering (DF), content-based filtering (CBF) and knowledge-based filtering (KBF). However, large amounts of data can produce recommendations that are limited in accuracy because of diversity and sparsity issues. In this paper, we propose a novel hybrid method that combines user–user CF with the attributes of DF to indicate the nearest users, and compare four classifiers against each other. This method has been developed through an investigation of ways to reduce the errors in rating predictions based on users’ past interactions, which leads to improved prediction accuracy in all four classification algorithms. We applied a feature combination method that improves the prediction accuracy and to test our approach, we ran an offline evaluation using the 1M MovieLens dataset, well-known evaluation metrics and comparisons between methods with the results validating our proposed method.


Sign in / Sign up

Export Citation Format

Share Document