Search results

Learning to rank algorithms inherently faces many challenges. The most important challenges could be listed as high-dimensionality of the training data, the dynamic nature of Web information resources and lack of click-through data. High dimensionality of the training data affects effectiveness and efficiency of learning algorithms. Besides, most of learning to rank benchmark datasets do not include click-through data as a very rich source of information about the search behavior of users while dealing with the ranked lists of search results. To deal with these limitations, this paper aims to introduce a novel learning to rank algorithm by using a set of complex click-through features in a reinforcement learning (RL) model. These features are calculated from the existing click-through information in the data set or even from data sets without any explicit click-through information.

The proposed ranking algorithm (QRC-Rank) applies RL techniques on a set of calculated click-through features. QRC-Rank is as a two-steps process. In the first step, Transformation phase, a compact benchmark data set is created which contains a set of click-through features. These feature are calculated from the original click-through information available in the data set and constitute a compact representation of click-through information. To find most effective click-through feature, a number of scenarios are investigated. The second phase is Model-Generation, in which a RL model is built to rank the documents. This model is created by applying temporal difference learning methods such as Q-Learning and SARSA.

The proposed learning to rank method, QRC-rank, is evaluated on WCL2R and LETOR4.0 data sets. Experimental results demonstrate that QRC-Rank outperforms the state-of-the-art learning to rank methods such as SVMRank, RankBoost, ListNet and AdaRank based on the precision and normalized discount cumulative gain evaluation criteria. The use of the click-through features calculated from the training data set is a major contributor to the performance of the system.

In this paper, we have demonstrated the viability of the proposed features that provide a compact representation for the click through data in a learning to rank application. These compact click-through features are calculated from the original features of the learning to rank benchmark data set. In addition, a Markov Decision Process model is proposed for the learning to rank problem using RL, including the sets of states, actions, rewarding strategy and the transition function.

The purpose of this paper is to investigate the applicability of meta‐learning to the problem of algorithm recommendation for gene expression data classification.

Meta‐learning was used to provide a preference order of machine learning algorithms, based on their expected performances. Two approaches were considered for such: k‐nearest neighbors and support vector machine‐based ranking methods. They were applied to a set of 49 publicly available microarray datasets. The evaluation of the methods followed standard procedures suggested in the meta‐learning literature.

Empirical evidences show that both ranking methods produce more interesting suggestions for gene expression data classification than the baseline method. Although the rankings are more accurate, a significant difference in the performances of the top classifiers was not observed.

As the experiments conducted in this paper suggest, the use of meta‐learning approaches can provide an efficient data driven way to select algorithms for gene expression data classification.

This paper reports contributions to the areas of meta‐learning and gene expression data analysis. Regarding the former, it supports the claim that meta‐learning can be suitably applied to problems of a specific domain, expanding its current practice. To the latter, it introduces a cost effective approach to better deal with classification tasks.

User feedback inferred from the user's search-time behavior could improve the learning to rank (L2R) algorithms. Click models (CMs) present probabilistic frameworks for describing and predicting the user's clicks during search sessions. Most of these CMs are based on common assumptions such as Attractiveness, Examination and User Satisfaction. CMs usually consider the Attractiveness and Examination as pre- and post-estimators of the actual relevance. They also assume that User Satisfaction is a function of the actual relevance. This paper extends the authors' previous work by building a reinforcement learning (RL) model to predict the relevance. The Attractiveness, Examination and User Satisfaction are estimated using a limited number of the features of the utilized benchmark data set and then they are incorporated in the construction of an RL agent. The proposed RL model learns to predict the relevance label of documents with respect to a given query more effectively than the baseline RL models for those data sets.

In this paper, User Satisfaction is used as an indication of the relevance level of a query to a document. User Satisfaction itself is estimated through Attractiveness and Examination, and in turn, Attractiveness and Examination are calculated by the random forest algorithm. In this process, only a small subset of top information retrieval (IR) features are used, which are selected based on their mean average precision and normalized discounted cumulative gain values. Based on the authors' observations, the multiplication of the Attractiveness and Examination values of a given query–document pair closely approximates the User Satisfaction and hence the relevance level. Besides, an RL model is designed in such a way that the current state of the RL agent is determined by discretization of the estimated Attractiveness and Examination values. In this way, each query–document pair would be mapped into a specific state based on its Attractiveness and Examination values. Then, based on the reward function, the RL agent would try to choose an action (relevance label) which maximizes the received reward in its current state. Using temporal difference (TD) learning algorithms, such as Q-learning and SARSA, the learning agent gradually learns to identify an appropriate relevance label in each state. The reward that is used in the RL agent is proportional to the difference between the User Satisfaction and the selected action.

Experimental results on MSLR-WEB10K and WCL2R benchmark data sets demonstrate that the proposed algorithm, named as SeaRank, outperforms baseline algorithms. Improvement is more noticeable in top-ranked results, which usually receive more attention from users.

This research provides a mapping from IR features to the CM features and thereafter utilizes these newly generated features to build an RL model. This RL model is proposed with the definition of the states, actions and reward function. By applying TD learning algorithms, such as the Q-learning and SARSA, within several learning episodes, the RL agent would be able to learn how to choose the most appropriate relevance label for a given pair of query–document.

This paper aims to learn a decision-maker’s (DM’s) decision model that is characterized in terms of the attitudinal character and the attributes weight vector, both of which are specific to the DM. The authors take the learning information in the form of the exemplary preferences, given by a DM. The learning approach is formalized by bringing together the recent research in the choice models and machine learning. The study is validated on a set of 12 benchmark data sets.

The study includes emerging preference learning algorithms.

Learning of a DM’s attitudinal choice model.

Preferences-based learning of a DM’s attitudinal decision model.

This study reviews the methods found in the literature for the ranking of authors, identifies the pros and cons of these methods, discusses and compares these methods. The purpose of this paper is to study is to find the challenges and future directions of ranking of academic objects, especially authors, for future researchers.

This study reviews the methods found in the literature for the ranking of authors, classifies them into subcategories by studying and analyzing their way of achieving the objectives, discusses and compares them. The data sets used in the literature and the evaluation measures applicable in the domain are also presented.

The survey identifies the challenges involved in the field of ranking of authors and future directions.

To the best of the knowledge, this is the first survey that studies the author ranking problem in detail and classifies them according to their key functionalities, features and way of achieving the objective according to the requirement of the problem.

Incorporating users’ behavior patterns could help in the ranking process. Different click models (CMs) are introduced to model the sophisticated search-time behavior of users among which commonly used the triple of attractiveness, examination and satisfaction. Inspired by this fact and considering the psychological definitions of these concepts, this paper aims to propose a novel learning to rank by redefining these concepts. The attractiveness and examination factors could be calculated using a limited subset of information retrieval (IR) features by the random forest algorithm, and then they are combined with each other to predicate the satisfaction factor which is considered as the relevance level.

The attractiveness and examination factors of a given document are usually considered as its perceived relevance and the fast scan of its snippet, respectively. Here, attractiveness and examination factors are regarded as the click-count and the investigation rate, respectively. Also, the satisfaction of a document is supposed to be the same as its relevance level for a given query. This idea is supported by the strong correlation between attractiveness-satisfaction and the examination-satisfaction. Applying random forest algorithm, the attractiveness and examination factors are calculated using a very limited set of the primitive features of query-document pairs. Then, by using the ordered weighted averaging operator, these factors are aggregated to estimate the satisfaction.

Experimental results on MSLR-WEB10K and WCL2R data sets show the superiority of this algorithm over the state-of-the-art ranking algorithms in terms of P@n and NDCG criteria. The enhancement is more noticeable in top-ranked items which are reviewed more by the users.

This paper proposes a novel learning to rank based on the redefinition of major building blocks of the CMs which are the attractiveness, examination and satisfactory. It proposes a method to use a very limited number of selected IR features to estimate the attractiveness and examination factors and then combines these factors to predicate the satisfactory which is regarded as the relevance level of a document with respect to a given query.

The personal learning environment driven approach to learning suggests a shift in emphasis from a teacher‐driven knowledge‐push to a learner‐driven knowledge‐pull learning model. One concern with knowledge‐pull approaches is knowledge overload. The concepts of collective intelligence and the Long Tail provide a potential solution to help learners cope with the problem of knowledge overload. The paper aims to address these issues.

Based on these concepts, the paper proposes a filtering mechanism that taps the collective intelligence to help learners find quality in the Long Tail, thus overcoming the problem of knowledge overload.

The paper presents theoretical, design, and implementation details of PLEM, a Web 2.0 driven service for personal learning management, which acts as a Long Tail aggregator and filter for learning.

The primary aim of PLEM is to harness the collective intelligence and leverage social filtering methods to rank and recommend learning entities.

The purpose of this paper is to explore the conceptual adequacy of the learning economy and its ability to describe the modern globalised economy. It is argued that unlike many fleeting catchwords and phrases found in economics, the learning economy represents a superior conceptual and heuristic starting point that reflects a new and emerging economic regime.

The paper examines those features of the learning economy which makes it a useful conceptualization and highlights some preconditions that are functional for its emergence. The paper then assesses the empirical validity of the learning economy and gauges its performance across 16 EU countries.

The learning economy represents a viable and useful concept in economics and the broader social sciences, which synthesizes recent attempts to depict what is new in the world economy into an internally coherent whole while overcoming previous shortcomings. It reflects a tangible reality that has taken hold most firmly in a small but significant part of the world, the Nordic countries of Northwestern Europe.

Because it has already emerged in some advanced countries, the learning economy offers a concrete exemplar for other countries to emulate. If one has to ask people and communities to sacrifice, save and invest for the future, it is more convincing to do so for a concrete and credible future that does exist than for some conjectural future.

The paper uses an epistemological perspective to analyse the concept of the learning economy as articulated by Bengt-Äke Lundvall.

Presently, searching the internet for learning material relevant to ones own interest continues to be a time‐consuming task. Systems that can suggest learning material (learning objects) to a learner would reduce time spent searching for material, and enable the learner to spend more time for actual learning. The purpose of this paper is to present a system of “hybrid search and delivery of learning objects to learners”.

This paper presents a system of “hybrid search and delivery of learning objects to learners” that combines the use of WordNet for semantic query expansion and an approach to personalized learning object delivery by suggesting relevant learning objects based on attributes specified in the learner's profile. The learning objects are related to the learner's attributes using the IEEE LOM and IMS LIP standards. The system includes a web crawler to collect learning objects from existing learning object repositories, such as NEEDS or SMETE.

The presented HSDLO system has the ability to accurately search and deliver learning objects of interest to a learner as well as adjust the learner's profile over time by evaluating the learner's preferences implicitly through the learning object selections.

Since real LOM's from SMETE are not much populated, the system is tested with a limited set of attributes. The system is evaluated using a test bench rather than real learners.

The paper proposes a combination of three search techniques in one system as well an architectural solution which can be used for other types of online search engines.

Social media platform, like Twitter, has increasingly become the mode of reporting civic issues owing to their vast and fast reachability. A tremendous amount of information on urban issues is shared every moment out of which some tweets may need immediate attention to save lives or avoid future disasters. Existing approaches are only limited to the identification of complaint tweets; however, its prioritization based on urgency is still unexplored. This study aims to decide the ranking of complaints based on its criticality derived using multiple parameters, like type of complaint, season, day or night, gender, holiday or working day, etc.

The approach proposes an ensemble of multi-class classification (MCC) and “two-level” multi-criteria decision-making (MCDM) algorithms, like AHP and TOPSIS, to evaluate the accurate ranking score of the tweet based on the severity of the issue. Initially, the MCC is applied to tweets to categorize the tweets into three categories, i.e. moderate, urgent and immediate. Further, the first level of MCDM algorithm decides the ranking within each complaint type, and the second level evaluates the ranking across all types. Integration of MCC and MCDM methods further helps to increase the accuracy of the result.

The paper discusses various parameters and investigates how their combination plays a significant role in deciding the priority of complaints. It successfully demonstrates that MCDM techniques are helpful in generating the ranking score of tweets based on various criteria.

This paper fulfills an identified need to prioritize the complaint tweet which helps the local government to take time-bound actions and save a life.