Ranking Support Vector Research Articles

Feedback Matching Framework for Semantic Interoperability of Product Data

There is a need to promote drastically increased levels of interoperability of product data across a broad spectrum of stakeholders, while ensuring that the semantics of product knowledge are preserved, and when necessary, translated. In order to achieve this, multiple methods have been proposed to determine semantic maps across concepts from different representations. Previous research has focused on developing different individual matching methods, i.e., ones that compute mapping based on a single matching measure. These efforts assume that some weighted combination can be used to obtain the overall maps. We analyze the problem of combination of multiple individual methods to determine requirements specific to product development and propose a solution approach called FEedback Matching Framework with Implicit Training (FEMFIT). FEMFIT provides the ability to combine the different matching approaches using ranking Support Vector Machine (ranking SVM). The method accounts for nonlinear relations between the individual matchers. It overcomes the need to explicitly train the algorithm before it is used, and further reduces the decision-making load on the domain expert by implicitly capturing the expert's decisions without requiring him to input real numbers on similarity. We apply FEMFIT to a subset of product constraints across a commercial system and the ISO standard. We observe that FEMIT demonstrates better accuracy (average correctness of the results) and stability (deviation from the average) in comparison with other existing combination methods commonly assumed to be valid in this domain.

Automatic Discovery of Association Orders between Name and Aliases from the Web using Anchor Texts-based Co-occurrences

Many celebrities and experts from various fields may have been referred by not only their personal names but also by their aliases on web. Aliases are very important in information retrieval to retrieve complete information about a personal name from the web, as some of the web pages of the person may also be referred by his aliases. The aliases for a personal name are extracted by previously proposed alias extraction method. In information retrieval, the web search engine automatically expands the search query on a person name by tagging his aliases for complete information retrieval thereby improving recall in relation detection task and achieving a significant mean reciprocal rank (MRR) of search engine. For the further substantial improvement on recall and MRR from the previously proposed methods, our proposed method will order the aliases based on their associations with the name using the definition of anchor texts-based co-occurrences between name and aliases in order to help the search engine tag the aliases according to the order of associations. The association orders will automatically be discovered by creating an anchor texts-based co-occurrence graph between name and aliases. Ranking support vector machine (SVM) will be used to create connections between name and aliases in the graph by performing ranking on anchor texts-based co-occurrence measures. The hop distances between nodes in the graph will lead to have the associations between name and aliases. The hop distances will be found by mining the graph. The proposed method will outperform previously proposed methods, achieving substantial growth on recall and MRR.

Learning to Rank with Bayesian Evidence Framework

The problem of ranking has recently gained attention in data learning. The goal ranking is to learn a real-valued ranking function that induces a ranking or ordering over an instance space. In this paper, we apply popular Bayesian techniques on ranking support vector machine. We propose a novel differentiable loss function called trigonometric loss function with the desirable characteristic of natural normalization in the likelihood function, and then follow standard Gaussian processes techniques to set up a Bayesian framework. In this framework, Bayesian inference is used to implement model adaptation, while keeping the merits of ranking SVM. Experimental results on data sets indicate the usefulness of this approach.

Predicting Author Blog Channels with High Value Future Posts for Monitoring

The phenomenal growth of social media, both in scale and importance, has created a unique opportunity to track information diffusion and the spread of influence, but can also make efficient tracking difficult. Given data streams representing blog posts on multiple blog channels and a focal query post on some topic of interest, our objective is to predict which of those channels are most likely to contain a future post that is relevant, or similar, to the focal query post. We denote this task as the future author prediction problem (FAPP). This problem has applications in information diffusion for brand monitoring and blog channel personalization and recommendation. We develop prediction methods inspired by (naive) information retrieval approaches that use historical posts in the blog channel for prediction. We also train a ranking support vector machine (SVM) to solve the problem. We evaluate our methods on an extensive social media dataset; despite the difficulty of the task, all methods perform reasonably well. Results show that ranking SVM prediction can exploit blog channel and diffusion characteristics to improve prediction accuracy. Moreover, it is surprisingly good for prediction in emerging topics and identifying inconsistent authors.

Ranking support vector machine for multiple kernels output combination in protein–protein interaction extraction from biomedical literature

Knowledge about protein-protein interactions (PPIs) unveils the molecular mechanisms of biological processes. However, the volume and content of published biomedical literature on protein interactions is expanding rapidly, making it increasingly difficult for interaction database curators to detect and curate protein interaction information manually. We present a multiple kernel learning-based approach for automatic PPI extraction from biomedical literature. The approach combines the following kernels: feature-based, tree, and graph and combines their output with Ranking support vector machine (SVM). Experimental evaluations show that the features in individual kernels are complementary and the kernel combined with Ranking SVM achieves better performance than those of the individual kernels, equal weight combination and optimal weight combination. Our approach can achieve state-of-the-art performance with respect to the comparable evaluations, with 64.88% F-score and 88.02% AUC on the AImed corpus.

Predicting Author Blog Channels with High Value Future Posts for Monitoring

The phenomenal growth of social media, both in scale and importance, has created a unique opportunity to track information diffusion and the spread of influence, but can also make efficient tracking difficult. Given data streams representing blog posts on multiple blog channels and a focal query post on some topic of interest, our objective is to predict which of those channels are most likely to contain a future post that is relevant, or similar, to the focal query post. We denote this task as the future author prediction problem (FAPP). This problem has applications in information diffusion for brand monitoring and blog channel personalization and recommendation. We develop prediction methods inspired by (naive) information retrieval approaches that use historical posts in the blog channel for prediction. We also train a ranking support vector machine (SVM) to solve the problem. We evaluate our methods on an extensive social media dataset; despite the difficulty of the task, all methods perform reasonably well. Results show that ranking SVM prediction can exploit blog channel and diffusion characteristics to improve prediction accuracy. Moreover, it is surprisingly good for prediction in emerging topics and identifying inconsistent authors.

Automatic Discovery of Personal Name Aliases from the Web

An individual is typically referred by numerous name aliases on the web. Accurate identification of aliases of a given person name is useful in various web related tasks such as information retrieval, sentiment analysis, personal name disambiguation, and relation extraction. We propose a method to extract aliases of a given personal name from the web. Given a personal name, the proposed method first extracts a set of candidate aliases. Second, we rank the extracted candidates according to the likelihood of a candidate being a correct alias of the given name. We propose a novel, automatically extracted lexical pattern-based approach to efficiently extract a large set of candidate aliases from snippets retrieved from a web search engine. We define numerous ranking scores to evaluate candidate aliases using three approaches: lexical pattern frequency, word co-occurrences in an anchor text graph, and page counts on the web. To construct a robust alias detection system, we integrate the different ranking scores into a single ranking function using ranking support vector machines. We evaluate the proposed method on three data sets: an English personal names data set, an English place names data set, and a Japanese personal names data set. The proposed method outperforms numerous baselines and previously proposed name alias extraction methods, achieving a statistically significant mean reciprocal rank (MRR) of 0.67. Experiments carried out using location names and Japanese personal names suggest the possibility of extending the proposed method to extract aliases for different types of named entities, and for different languages. Moreover, the aliases extracted using the proposed method are successfully utilized in an information retrieval task and improve recall by 20 percent in a relation-detection task.

Training linear ranking SVMs in linearithmic time using red–black trees

We introduce an efficient method for training the linear ranking support vector machine. The method combines cutting plane optimization with red–black tree based approach to subgradient calculations, and has O( ms + mlog ( m)) time complexity, where m is the number of training examples, and s the average number of non-zero features per example. Best previously known training algorithms achieve the same efficiency only for restricted special cases, whereas the proposed approach allows any real valued utility scores in the training data. Experiments demonstrate the superior scalability of the proposed approach, when compared to the fastest existing RankSVM implementations.

Brain–Computer Evolutionary Multiobjective Optimization: A Genetic Algorithm Adapting to the Decision Maker

The centrality of the decision maker (DM) is widely recognized in the multiple criteria decision-making community. This translates into emphasis on seamless human-computer interaction, and adaptation of the solution technique to the knowledge which is progressively acquired from the DM. This paper adopts the methodology of reactive search optimization (RSO) for evolutionary interactive multiobjective optimization. RSO follows to the paradigm of “learning while optimizing,” through the use of online machine learning techniques as an integral part of a self-tuning optimization scheme. User judgments of couples of solutions are used to build robust incremental models of the user utility function, with the objective to reduce the cognitive burden required from the DM to identify a satisficing solution. The technique of support vector ranking is used together with a k-fold cross-validation procedure to select the best kernel for the problem at hand, during the utility function training procedure. Experimental results are presented for a series of benchmark problems.

An unsupervised self-organizing learning with support vector ranking for imbalanced datasets

The aim of computational learning algorithm is to establish grounds that work for any types of data, once and for all. However, majority of the classifiers have their base from balanced datasets. This paper discusses the issues related to imbalanced data distribution problem and the common strategy to deal with imbalance datasets. We propose a model capable of handling imbalance datasets well in which other typical classifiers fail to do so. The model adopted a derivation of support vector machines in selecting variables so that the problem of imbalanced data distribution can be relaxed. Then, we used an Emergent Self-Organizing Map (ESOM) to cluster the ranker features so as to provide clusters for unsupervised classification. This work progresses by examining the efficiency of the model in evaluating imbalanced datasets. Experimental results show that the criterion based on weight vector derivative achieves good results and performs consistently well over imbalance datasets. In general, our approach outperforms other classification methods which are unable to handle the imbalanced data distribution in the testing datasets.

A Statistical Diagnosis Approach for Analyzing Design–Silicon Timing Mismatch

Explaining the mismatch between predicted timing behavior from modeling and simulation, and the observed timing behavior measured on silicon chips can be very challenging. Given a list of potential sources, the mismatch can be the aggregate result caused by some of them both individually and collectively, resulting in a very large search space. Furthermore, observed data are always corrupted by some unknown statistical random noises. In this paper, we examine how trying to explain the mismatch observed on silicon can be classified as an ill-posed problem, where ill posed means that the solution may not be unique or stable. Thus, a small change in the observed response can have a large change in the predicted solution. To solve ill-posed problems, a statistical learning theory uses a principle called regularization. This paper proposes using a statistical learning method called support vector (SV) analysis to statistically analyze all known sources of uncertainty with the objective to rank which sources contribute the most to the observed mismatch. Experimental results are presented under different error assumption models to compare two kinds of SV ranking approaches to four other ranking approaches, where some use the idea of regularization and others do not. This paper is concluded by showing a self cross-validation approach to validate the ranking results when there is no true ranking available, as the case with actual silicon.

Ordinal Regression for Information Retrieval

This letter presents a new discriminative model for Information Retrieval (IR), referred to as Ordinal Regression Model (ORM). ORM is different from most existing models in that it views IR as ordinal regression problem (i.e. ranking problem) instead of binary classification. It is noted that the task of IR is to rank documents according to the user information needed, so IR can be viewed as ordinal regression problem. Two parameter learning algorithms for ORM are presented. One is a perceptron-based algorithm. The other is the ranking Support Vector Machine (SVM). The effectiveness of the proposed approach has been evaluated on the task of ad hoc retrieval using three English Text REtrieval Conference (TREC) sets and two Chinese TREC sets. Results show that ORM significantly outperforms the state-of-the-art language model approaches and OKAPI system in all test sets; and it is more appropriate to view IR as ordinal regression other than binary classification.

A multi-ranker model for adaptive XML searching

The evolution of computing technology suggests that it has become more feasible to offer access to Web information in a ubiquitous way, through various kinds of interaction devices such as PCs, laptops, palmtops, and so on. As XML has become a de-facto standard for exchanging Web data, an interesting and practical research problem is the development of models and techniques to satisfy various needs and preferences in searching XML data. In this paper, we employ a list of simple XML tagged keywords as a vehicle for searching XML fragments in a collection of XML documents. In order to deal with the diversified nature of XML documents as well as user preferences, we propose a novel multi-ranker model (MRM), which is able to abstract a spectrum of important XML properties and adapt the features to different XML search needs. The MRM is composed of three ranking levels. The lowest level consists of two categories of similarity and granularity features. At the intermediate level, we define four tailored XML rankers (XRs), which consist of different lower level features and have different strengths in searching XML fragments. The XRs are trained via a learning mechanism called the Ranking Support Vector Machine in a voting Spy Naïve Bayes framework (RSSF). The RSSF takes as input a set of labeled fragments and feature vectors and generates as output Adaptive Rankers (ARs) in the learning process. The ARs are defined over the XRs and generated at the top level of the MRM. We show empirically that the RSSF is able to improve the MRM significantly in the learning process that needs only a small set of training XML fragments. We demonstrate that the trained MRM is able to bring out the strengths of the XRs in order to adapt different preferences and queries.

A co-training framework for searching XML documents

In this paper, we study the use of XML tagged keywords (or simply key-tags) to search an XML fragment in a collection of XML documents. We present techniques that are able to employ users’ evaluations as feedback and then to generate an adaptive ranked list of XML fragments as the search results. First, we extend the vector space model as a basis to search XML fragments. The model examines the relevance between the imposed key-tags and identified fragments in XML documents, and determines the ranked result as an output. Second, in order to deal with the diversified nature of XML documents, we present four XML Rankers (XRs), which have different strengths in terms of similarity, granularity, and ranking features. The XRs are specially tailored to diversified XML documents. We then evaluate the XML search effectiveness and quality for each tailored XR and propose a meta-XML ranker (MXR) comprising the four XRs. The MXR is trained via a machine learning training scheme, which we term the ranking support vector machine (RSVM) in a co-training framework (RSCF). The RSCF takes as input two sets of labelled fragments and feature vectors and then generates as output adaptive rankers in a learning process. We show empirically that, with only a small set of training XML fragments, the RSCF is able to improve after a few iterations in the learning process. Finally, we demonstrate that the RSCF-based MXR is able to bring out the strengths of the underlying XRs in order to adapt the users’ perspectives on the returned search results. By using a set of key-tag queries on a variety of XML documents, we show that the precision of the result of the RSCF-based MXR is effective.