اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدFast and accurate annotation of short texts with Wikipedia pages
38
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
We address the problem of cross-referencing text fragments with Wikipedia pages, in a way that synonymy and polysemy issues are resolved accurately and efficiently. We take inspiration from a recent flow of work [Cucerzan 2007, Mihalcea and Csomai 2007, Milne and Witten 2008, Chakrabarti et al 2009], and extend their scenario from the annotation of long documents to the annotation of short texts, such as snippets of search-engine results, tweets, news, blogs, etc.. These short and poorly composed texts pose new challenges in terms of efficiency and effectiveness of the annotation process, that we address by designing and engineering TAGME, the first system that performs an accurate and on-the-fly annotation of these short textual fragments. A large set of experiments shows that TAGME outperforms state-of-the-art algorithms when they are adapted to work on short texts and it results fast and competitive on long texts.
قيم البحث
اقرأ أيضاً
We study the problem of entity salience by proposing the design and implementation of SWAT, a system that identifies the salient Wikipedia entities occurring in an input document. SWAT consists of several modules that are able to detect and classify
on-the-fly Wikipedia entities as salient or not, based on a large number of syntactic, semantic and latent features properly extracted via a supervised process which has been trained over millions of examples drawn from the New York Times corpus. The validation process is performed through a large experimental assessment, eventually showing that SWAT improves known solutions over all publicly available datasets. We release SWAT via an API that we describe and comment in the paper in order to ease its use in other software.
For providing quick and accurate search results, a search engine maintains a local snapshot of the entire web. And, to keep this local cache fresh, it employs a crawler for tracking changes across various web pages. It would have been ideal if the cr
awler managed to update the local snapshot as soon as a page changed on the web. However, finite bandwidth availability and server restrictions mean that there is a bound on how frequently the different pages can be crawled. This then brings forth the following optimisation problem: maximise the freshness of the local cache subject to the crawling frequency being within the prescribed bounds. Recently, tractable algorithms have been proposed to solve this optimisation problem under different cost criteria. However, these assume the knowledge of exact page change rates, which is unrealistic in practice. We address this issue here. Specifically, we provide three novel schemes for online estimation of page change rates. All these schemes only need partial information about the page change process, i.e., they only need to know if the page has changed or not since the last crawl instance. Our first scheme is based on the law of large numbers, the second on the theory of stochastic approximation, while the third is an extension of the second and involves an additional momentum term. For all of these schemes, we prove convergence and, also, provide their convergence rates. As far as we know, the results concerning the third estimator is quite novel. Specifically, this is the first convergence type result for a stochastic approximation algorithm with momentum. Finally, we provide some numerical experiments (on real as well as synthetic data) to compare the performance of our proposed estimators with the existing ones (e.g., MLE).
We argue that relationships between Web pages are functions of the users intent. We identify a class of Web tasks - information-gathering - that can be facilitated by a search engine that provides links to pages which are related to the page the user
is currently viewing. We define three kinds of intentional relationships that correspond to whether the user is a) seeking sources of information, b) reading pages which provide information, or c) surfing through pages as part of an extended information-gathering process. We show that these three relationships can be productively mined using a combination of textual and link information and provide three scoring mechanisms that correspond to them: {em SeekRel}, {em FactRel} and {em SurfRel}. These scoring mechanisms incorporate both textual and link information. We build a set of capacitated subnetworks - each corresponding to a particular keyword - that mirror the interconnection structure of the World Wide Web. The scores are computed by computing flows on these subnetworks. The capacities of the links are derived from the {em hub} and {em authority} values of the nodes they connect, following the work of Kleinberg (1998) on assigning authority to pages in hyperlinked environments. We evaluated our scoring mechanism by running experiments on four data sets taken from the Web. We present user evaluations of the relevance of the top results returned by our scoring mechanisms and compare those to the top results returned by Googles Similar Pages feature, and the {em Companion} algorithm proposed by Dean and Henzinger (1999).
Blog is becoming an increasingly popular media for information publishing. Besides the main content, most of blog pages nowadays also contain noisy information such as advertisements etc. Removing these unrelated elements can improves user experience
, but also can better adapt the content to various devices such as mobile phones. Though template-based extractors are highly accurate, they may incur expensive cost in that a large number of template need to be developed and they will fail once the template is updated. To address these issues, we present a novel template-independent content extractor for blog pages. First, we convert a blog page into a DOM-Tree, where all elements including the title and body blocks in a page correspond to subtrees. Then we construct subtree candidate set for the title and the body blocks respectively, and extract both spatial and content features for elements contained in the subtree. SVM classifiers for the title and the body blocks are trained using these features. Finally, the classifiers are used to extract the main content from blog pages. We test our extractor on 2,250 blog pages crawled from nine blog sites with obviously different styles and templates. Experimental results verify the effectiveness of our extractor.
In multi-label text classification, each textual document can be assigned with one or more labels. Due to this nature, the multi-label text classification task is often considered to be more challenging compared to the binary or multi-class text clas
sification problems. As an important task with broad applications in biomedicine such as assigning diagnosis codes, a number of different computational methods (e.g. training and combining binary classifiers for each label) have been proposed in recent years. However, many suffered from modest accuracy and efficiency, with only limited success in practical use. We propose ML-Net, a novel deep learning framework, for multi-label classification of biomedical texts. As an end-to-end system, ML-Net combines a label prediction network with an automated label count prediction mechanism to output an optimal set of labels by leveraging both predicted confidence score of each label and the contextual information in the target document. We evaluate ML-Net on three independent, publicly-available corpora in two kinds of text genres: biomedical literature and clinical notes. For evaluation, example-based measures such as precision, recall and f-measure are used. ML-Net is compared with several competitive machine learning baseline models. Our benchmarking results show that ML-Net compares favorably to the state-of-the-art methods in multi-label classification of biomedical texts. ML-NET is also shown to be robust when evaluated on different text genres in biomedicine. Unlike traditional machine learning methods, ML-Net does not require human efforts in feature engineering and is highly efficient and scalable approach to tasks with a large set of labels (no need to build individual classifiers for each separate label). Finally, ML-NET is able to dynamically estimate the label count based on the document context in a more systematic and accurate manner.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
