Short text classification using very few words

Sun, Aixin

doi:10.1145/2348283.2348511

Cited by 83 publications

(37 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate the performance of the above approach, we use the Web snippet dataset used in (Phan et al, 2008;Chen et al, 2011;Sun, 2012 Table 2: The number of unseen words to query the web search engine (Google) and selected the top 20 or 30 snippets from the search results. Different phrases for the training and test data were used to make sure that test data were difficult to classify (Phan et al, 2008).…”

Section: Dataset and Evaluation Metricsmentioning

confidence: 99%

Distributional Representations of Words for Short Text Classification

et al. 2015

Proceedings of the 1st Workshop on Vector Space Modeling for Natural Language Processing

View full text Add to dashboard Cite

Traditional supervised learning approaches to common NLP tasks depend heavily on manual annotation, which is labor intensive and time consuming, and often suffer from data sparseness. In this paper we show how to mitigate the problems in short text classification (STC) through word embeddings -distributional representations of words learned from large unlabeled data. The word embeddings are trained from the entire English Wikipedia text. We assume that a short text document is a specific sample of one distribution in a Bayesian framework. A Gaussian process approach is used to model the distribution of words. The task of classification becomes a simple problem of selecting the most probable Gaussian distribution. This approach is compared with those based on the classical maximum entropy (MaxEnt) model and the Latent Dirichlet Allocation (LDA) approach. Our approach achieved better performance and also showed advantages in dealing with unseen words.

show abstract

Section: Dataset and Evaluation Metricsmentioning

confidence: 99%

Distributional Representations of Words for Short Text Classification

et al. 2015

Proceedings of the 1st Workshop on Vector Space Modeling for Natural Language Processing

View full text Add to dashboard Cite

show abstract

“…Based on traditional texts such as KNN [3], Bayesian classification [4], decision tree [5], SVM [6], and maximum entropy [7], previous classification methods have achieved good results in text classification. However, these methods require enough co-occurrence information of word frequency in texts, and it is not effective when applying to short text classification.…”

Section: Related Workmentioning

confidence: 99%

Research of specific field ultra-short text classification based on collaborative filtering algorithm

Yang

2018

MATEC Web Conf.

View full text Add to dashboard Cite

Abstract. In some specific fields, there are a lot of ultra-short texts that need to be categorized. This paper proposes an ultra-short text classification method based on collaborative filtering algorithm aiming at the problems such as short text content, short length, sparse features, and large number of categories in certain fields. First, converting ultra-short text into word frequency vector by doing Chinese word segmentation and calculating word frequency; Secondly, combining relevant data in specific fields, defining the ultra-short texts as users, categories as items, and then constructing a user-item recommendation matrix. Finally, calculating text similarity by using cosine similarity method and obtaining the classification results. The experimental results show that the proposed method can well solve the problem of classification of ultra-short texts in specific fields, and the average accuracy is 9.19% and 3.81% higher than vector space model and topic similarity method respectively.

show abstract

“…Likely, Chen et al, moved forward along this direction and proposed a method to leverage topics at multiple granularities [28]. In other ways, a search and vote strategy with search results was used in labeling the query candidates in Sun's paper [29]. Meng et al, presents an effective algorithm for semantic similarity metric of word pairs [30] and a new similar queries metric for query suggestions, then measure similar queries from the level of semantic level [31].…”

Section: Short Text Processingmentioning

confidence: 99%