arXiv subjects

Joeran Beel

Publications and source records attributed to Joeran Beel.

47 records · Page 3Linked to original sources

arXiv2018

Implementing Neural Turing Machines

Neural Turing Machines (NTMs) are an instance of Memory Augmented Neural Networks, a new class of recurrent neural networks which decouple computation from memory by introducing an external memory unit. NTMs have demonstrated superior performance over Long Short-Term Memory Cells in several sequence learning tasks. A number of open source implementations of NTMs exist but are unstable during training and/or fail to replicate the reported performance of NTMs. This paper presents the details of our successful implementation of a NTM. Our implementation learns to solve three sequential learning tasks from the original NTM paper. We find that the choice of memory contents initialization scheme is crucial in successfully implementing a NTM. Networks with memory contents initialized to small constant values converge on average 2 times faster than the next best memory contents initialization scheme.

cs.LG↗

arXiv2018

Machine Learning vs. Rules and Out-of-the-Box vs. Retrained: An Evaluation of Open-Source Bibliographic Reference and Citation Parsers

Bibliographic reference parsing refers to extracting machine-readable metadata, such as the names of the authors, the title, or journal name, from bibliographic reference strings. Many approaches to this problem have been proposed so far, including regular expressions, knowledge bases and supervised machine learning. Many open source reference parsers based on various algorithms are also available. In this paper, we apply, evaluate and compare ten reference parsing tools in a specific business use case. The tools are Anystyle-Parser, Biblio, CERMINE, Citation, Citation-Parser, GROBID, ParsCit, PDFSSA4MET, Reference Tagger and Science Parse, and we compare them in both their out-of-the-box versions and versions tuned to the project-specific data. According to our evaluation, the best performing out-of-the-box tool is GROBID (F1 0.89), followed by CERMINE (F1 0.83) and ParsCit (F1 0.75). We also found that even though machine learning-based tools and tools based on rules or regular expressions achieve on average similar precision (0.77 for ML-based tools vs. 0.76 for non-ML-based tools), applying machine learning-based tools results in a recall three times higher than in the case of non-ML-based tools (0.66 vs. 0.22). Our study also confirms that tuning the models to the task-specific data results in the increase in the quality. The retrained versions of reference parsers are in all cases better than their out-of-the-box counterparts; for GROBID F1 increased by 3% (0.92 vs. 0.89), for CERMINE by 11% (0.92 vs. 0.83), and for ParsCit by 16% (0.87 vs. 0.75).

cs.DL↗

arXiv2018

A Study of Position Bias in Digital Library Recommender Systems

"Position bias" describes the tendency of users to interact with items on top of a list with higher probability than with items at a lower position in the list, regardless of the items' actual relevance. In the domain of recommender systems, particularly recommender systems in digital libraries, position bias has received little attention. We conduct a study in a real-world recommender system that delivered ten million related-article recommendations to the users of the digital library Sowiport, and the reference manager JabRef. Recommendations were randomly chosen to be shuffled or non-shuffled, and we compared click-through rate (CTR) for each rank of the recommendations. According to our analysis, the CTR for the highest rank in the case of Sowiport is 53% higher than expected in a hypothetical non-biased situation (0.189% vs. 0.123%). Similarly, in the case of Jabref the highest rank received a CTR of 1.276%, which is 87% higher than expected (0.683%). A chi-squared test confirms the strong relationship between the rank of the recommendation shown to the user and whether the user decided to click it (p < 0.01 for both Jabref and Sowiport). Our study confirms the findings from other domains, that recommendations in the top positions are more often clicked, regardless of their actual relevance.

cs.DL↗

arXiv2018

A Method for Discovering and Extracting Author Contributions Information from Scientific Biomedical Publications

Creating scientific publications is a complex process, typically composed of a number of different activities, such as designing the experiments, data preparation, programming software and writing and editing the manuscript. The information about the contributions of individual authors of a paper is important in the context of assessing authors' scientific achievements. Some publications in biomedical disciplines contain a description of authors' roles in the form of a short section written in natural language, typically entitled "Authors' contributions". In this paper, we present an analysis of roles commonly appearing in the content of these sections, and propose an algorithm for automatic extraction of authors' roles from natural language text in scientific publications. During the first part of the study, we used clustering techniques, as well as Open Information Extraction (OpenIE), to semi-automatically discover the most popular roles within a corpus of 2,000 contributions sections obtained from PubMed Central resources. The roles discovered by our approach include: experimenting (1,743 instances, 17% of the entire role set within the corpus), analysis (1,343, 16%), study design (1,132, 13%), interpretation (879, 10%), conceptualization (865, 10%), paper reading (823, 10%), paper writing (724, 8%), paper review (501, 6%), paper drafting (351, 4%), coordination (319, 4%), data collection (76, 1%), paper review (41, 0.5%) and literature review (41, 0.5%). Discovered roles were then used to automatically build a training set for the supervised role extractor, based on Naive Bayes algorithm. According to the evaluation we performed, the proposed role extraction algorithm is able to extract the roles from the text with precision 0.71, recall 0.49 and F1 0.58.

cs.DL↗

arXiv2017

It's Time to Consider "Time" when Evaluating Recommender-System Algorithms [Proposal]

In this position paper, we question the current practice of calculating evaluation metrics for recommender systems as single numbers (e.g. precision p=.28 or mean absolute error MAE = 1.21). We argue that single numbers express only average effectiveness over a usually rather long period (e.g. a year or even longer), which provides only a vague and static view of the data. We propose that recommender-system researchers should instead calculate metrics for time-series such as weeks or months, and plot the results in e.g. a line chart. This way, results show how algorithms' effectiveness develops over time, and hence the results allow drawing more meaningful conclusions about how an algorithm will perform in the future. In this paper, we explain our reasoning, provide an example to illustrate our reasoning and present suggestions for what the community should do next.

cs.IR↗

arXiv2017

RARD: The Related-Article Recommendation Dataset

Recommender-system datasets are used for recommender-system evaluations, training machine-learning algorithms, and exploring user behavior. While there are many datasets for recommender systems in the domains of movies, books, and music, there are rather few datasets from research-paper recommender systems. In this paper, we introduce RARD, the Related-Article Recommendation Dataset, from the digital library Sowiport and the recommendation-as-a-service provider Mr. DLib. The dataset contains information about 57.4 million recommendations that were displayed to the users of Sowiport. Information includes details on which recommendation approaches were used (e.g. content-based filtering, stereotype, most popular), what types of features were used in content based filtering (simple terms vs. keyphrases), where the features were extracted from (title or abstract), and the time when recommendations were delivered and clicked. In addition, the dataset contains an implicit item-item rating matrix that was created based on the recommendation click logs. RARD enables researchers to train machine learning algorithms for research-paper recommendations, perform offline evaluations, and do research on data from Mr. DLib's recommender system, without implementing a recommender system themselves. In the field of scientific recommender systems, our dataset is unique. To the best of our knowledge, there is no dataset with more (implicit) ratings available, and that many variations of recommendation algorithms. The dataset is available at http://data.mr-dlib.org, and published under the Creative Commons Attribution 3.0 Unported (CC-BY) license.

cs.IR↗

arXiv2017

Mr. DLib: Recommendations-as-a-Service (RaaS) for Academia

Only few digital libraries and reference managers offer recommender systems, although such systems could assist users facing information overload. In this paper, we introduce Mr. DLib's recommendations-as-a-service, which allows third parties to easily integrate a recommender system into their products. We explain the recommender approaches implemented in Mr. DLib (content-based filtering among others), and present details on 57 million recommendations, which Mr. DLib delivered to its partner GESIS Sowiport. Finally, we outline our plans for future development, including integration into JabRef, establishing a living lab, and providing personalized recommendations.

cs.IR↗

arXiv2017

Exploring Choice Overload in Related-Article Recommendations in Digital Libraries

We investigate the problem of choice overload - the difficulty of making a decision when faced with many options - when displaying related-article recommendations in digital libraries. So far, research regarding to how many items should be displayed has mostly been done in the fields of media recommendations and search engines. We analyze the number of recommendations in current digital libraries. When browsing fullscreen with a laptop or desktop PC, all display a fixed number of recommendations. 72% display three, four, or five recommendations, none display more than ten. We provide results from an empirical evaluation conducted with GESIS' digital library Sowiport, with recommendations delivered by recommendations-as-a-service provider Mr. DLib. We use click-through rate as a measure of recommendation effectiveness based on 3.4 million delivered recommendations. Our results show lower click-through rates for higher numbers of recommendations and twice as many clicked recommendations when displaying ten instead of one related-articles. Our results indicate that users might quickly feel overloaded by choice.

cs.IR↗

arXiv2017

Real-World Recommender Systems for Academia: The Pain and Gain in Building, Operating, and Researching them [Long Version]

Research on recommender systems is a challenging task, as is building and operating such systems. Major challenges include non-reproducible research results, dealing with noisy data, and answering many questions such as how many recommendations to display, how often, and, of course, how to generate recommendations most effectively. In the past six years, we built three research-article recommender systems for digital libraries and reference managers, and conducted research on these systems. In this paper, we share some experiences we made during that time. Among others, we discuss the required skills to build recommender systems, and why the literature provides little help in identifying promising recommendation approaches. We explain the challenge in creating a randomization engine to run A/B tests, and how low data quality impacts the calculation of bibliometrics. We further discuss why several of our experiments delivered disappointing results, and provide statistics on how many researchers showed interest in our recommendation dataset.

cs.IR↗

arXiv2017

Towards Effective Research-Paper Recommender Systems and User Modeling based on Mind Maps

While user-modeling and recommender systems successfully utilize items like emails, news, and movies, they widely neglect mind-maps as a source for user modeling. We consider this a serious shortcoming since we assume user modeling based on mind maps to be equally effective as user modeling based on other items. Hence, millions of mind-mapping users could benefit from user-modeling applications such as recommender systems. The objective of this doctoral thesis is to develop an effective user-modeling approach based on mind maps. To achieve this objective, we integrate a recommender system in our mind-mapping and reference-management software Docear. The recommender system builds user models based on the mind maps, and recommends research papers based on the user models. As part of our research, we identify several variables relating to mind-map-based user modeling, and evaluate the variables' impact on user-modeling effectiveness with an offline evaluation, a user study, and an online evaluation based on 430,893 recommendations displayed to 4,700 users. We find, among others, that the number of analyzed nodes, modification time, visibility of nodes, relations between nodes, and number of children and siblings of a node affect the effectiveness of user modeling. When all variables are combined in a favorable way, this novel approach achieves click-through rates of 7.20%, which is nearly twice as effective as the best baseline. In addition, we show that user modeling based on mind maps performs about as well as user modeling based on other items, namely the research articles users downloaded or cited. Our findings let us to conclude that user modeling based on mind maps is a promising research field, and that developers of mind-mapping applications should integrate recommender systems into their applications. Such systems could create additional value for millions of mind-mapping users.

cs.IR↗

arXiv2017

Apache Lucene as Content-Based-Filtering Recommender System: 3 Lessons Learned

For the past few years, we used Apache Lucene as recommendation frame-work in our scholarly-literature recommender system of the reference-management software Docear. In this paper, we share three lessons learned from our work with Lucene. First, recommendations with relevance scores below 0.025 tend to have significantly lower click-through rates than recommendations with relevance scores above 0.025. Second, by picking ten recommendations randomly from Lucene's top50 search results, click-through rate decreased by 15%, compared to recommending the top10 results. Third, the number of returned search results tend to predict how high click-through rates will be: when Lucene returns less than 1,000 search results, click-through rates tend to be around half as high as if 1,000+ results are returned.

cs.IR↗