Approximating Multi-class Text Classification Via Automatic Generation of Training Examples

Geraci, Filippo; Papini, Tiziano

doi:10.1007/978-3-319-77116-8_44

Cited by 3 publications

(4 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As expected, home pages are very informative containing most of the base information about the website. In particular, leveraging only on home pages it is possible to derive some structural information about the underneath technologies [17], accomplish large scale usability tests [18], [19] and perform web classification [20].…”

Section: Crawling Depthmentioning

confidence: 99%

“…In [20] we faced the problem of using a small description of a class (as small as a weighted list of keywords) to automatically derive from the Web a list of pages that are more likely to be good examples to train a classifier. As most semi-supervised methods, our approach attempts to balance the use of human resources with the decrease of classification accuracy.…”

Section: Creation Of the Training Setmentioning

confidence: 99%

“…In this section we sketch out our pipeline leaving details to the original work in [20]. Both the plethora of informal guides on how to choose a domain name and the scientific literature on the psychological effects of brand names agree that a successful brand should consist in the juxtaposition of few words.…”

Section: A Semi Supervised Pipeline For the Extraction Of Training Exmentioning

confidence: 99%

“…In particular, weakness is measured as the inverse degree of membership with the labelling class, while ambiguity is evaluated as the number of classes such that the degree of membership exceeds a pre-defined threshold. Experiments in [20] over 20,016 webpages belonging to 9 non-overlapping categories show that 32.36% of the URLs succeed to be tokenized and 26.33% of the pages are labeled with an accuracy of 85.8%.…”

Section: A Semi Supervised Pipeline For the Extraction Of Training Exmentioning

confidence: 99%

See 3 more Smart Citations

Web Crawling and Processing with Limited Resources for Business Intelligence and Analytics Applications

Genovese¹,

Geraci²

2018

JSW

Self Cite

View full text Add to dashboard Cite

Business intelligence (BI) is the activity of extracting strategic information from big data. The benefits of this activity for enterprises span from the reduction of the operative costs due to a more sensible internal organization to a more productive and aware decision process. To be effective, BI relies heavily on the availability of a huge amount of (possibly high-quality) data. The steady decrease of costs for acquiring, storing and analyzing large knowledge bases has motivated big companies to invest in BI technologies. Until now, instead, SMEs (Small and Medium-sized Companies) are excluded from the benefits of BI because of their limited budget and resources. In this paper we show that a satisfactory BI activity is possible even in presence of a small budget. Our ultimate goal is not necessarily that of proposing novel solutions but providing the practitioners with a sort of hitchhiker's guide to a cost-effective web-based BI. In particular, we discuss how the Web can be used as a cheap yet reliable source of information where crawling, data cleaning and classification can be achieved using a limited amount of CPU, storage space and bandwidth.

show abstract

Section: Crawling Depthmentioning

confidence: 99%

Section: Creation Of the Training Setmentioning

confidence: 99%

Section: A Semi Supervised Pipeline For the Extraction Of Training Exmentioning

confidence: 99%

Section: A Semi Supervised Pipeline For the Extraction Of Training Exmentioning

confidence: 99%

See 2 more Smart Citations

Web Crawling and Processing with Limited Resources for Business Intelligence and Analytics Applications

Genovese¹,

Geraci²

2018

JSW

Self Cite

View full text Add to dashboard Cite

show abstract

BiGBERT: Classifying Educational Web Resources for Kindergarten-12$$^{th}$$ Grades

Allen

Downs

Shukla

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A review of semi-supervised learning for text classification

Duarte

Berton

2023

Artif Intell Rev

View full text Add to dashboard Cite

A huge amount of data is generated daily leading to big data challenges. One of them is related to text mining, especially text classification. To perform this task we usually need a large set of labeled data that can be expensive, time-consuming, or difficult to be obtained. Considering this scenario semi-supervised learning (SSL), the branch of machine learning concerned with using labeled and unlabeled data has expanded in volume and scope. Since no recent survey exists to overview how SSL has been used in text classification, we aim to fill this gap and present an up-to-date review of SSL for text classification. We retrieve 1794 works from the last 5 years from IEEE Xplore, ACM Digital Library, Science Direct, and Springer. Then, 157 articles were selected to be included in this review. We present the application domain, datasets, and languages employed in the works. The text representations and machine learning algorithms. We also summarize and organize the works following a recent taxonomy of SSL. We analyze the percentage of labeled data used, the evaluation metrics, and obtained results. Lastly, we present some limitations and future trends in the area. We aim to provide researchers and practitioners with an outline of the area as well as useful information for their current research.

show abstract

Approximating Multi-class Text Classification Via Automatic Generation of Training Examples

Cited by 3 publications

References 36 publications

Web Crawling and Processing with Limited Resources for Business Intelligence and Analytics Applications

Web Crawling and Processing with Limited Resources for Business Intelligence and Analytics Applications

BiGBERT: Classifying Educational Web Resources for Kindergarten-12$$^{th}$$ Grades

A review of semi-supervised learning for text classification

Contact Info

Product

Resources

About