A Fast Approach for Semantic Similar Short Texts Retrieval

Gu, Yu; Yang, Zhenglu; Zhou, Junsheng; Qu, Weiguang; Wei, Jinmao; Shi, Xingtian

doi:10.18653/v1/p16-2015

Cited by 5 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead, BewQA solves distinctive challenges in business servicing domains where the answer documents are semi-structured, more dynamic and low text-density. Beyond QA, another thread of research related to BewQA is short text retrieval [16]. These systems propose a fast approach to access a small subset of short text candidates, but require building appropriate indices offline.…”

Section: Related Workmentioning

confidence: 99%

Bew: Towards Answering Business-Entity-Related Web Questions

Cao,

Riva,

Balasubramanian

et al. 2020

Preprint

View full text Add to dashboard Cite

We present BewQA, a system specifically designed to answer a class of questions that we call Bew questions. Bew questions are related to businesses/services such as restaurants, hotels, and movie theaters; for example, "Until what time is happy hour?". These questions are challenging to answer because the answers are found in opendomain Web, are present in short sentences without surrounding context, and are dynamic since the webpage information can be updated frequently. Under these conditions, existing QA systems perform poorly. We present a practical approach, called BewQA, that can answer Bew queries by mining a template of the businessrelated webpages and using the template to guide the search. We show how we can extract the template automatically by leveraging aggregator websites that aggregate information about business entities in a domain (e.g., restaurants). We answer a given question by identifying the section from the extracted template that is most likely to contain the answer. By doing so we can extract the answers even when the answer span does not have sufficient context. Importantly, BewQA does not require any training. We crowdsource a new dataset of 1066 Bew questions and ground-truth answers in the restaurant domain. Compared to state-of-the-art QA models, BewQA has a 27 percent point improvement in F1 score. Compared to a commercial search engine, BewQA answered correctly 29% more Bew questions.

show abstract

Section: Related Workmentioning

confidence: 99%

Bew: Towards Answering Business-Entity-Related Web Questions

Cao,

Riva,

Balasubramanian

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…The similarity analysis has been done between queries [6], documents [1], text snippets [7,8], short segment [6,9], tweets [10,11] or question answer (QA) [12,13]. Most of the preceding works on semantic similarity or combining the semantic and lexical model relies on additional information derived from large corpora, dictionary [14,15] or background knowledge such as WordNet or ConceptNet [16,17]. Thus these work highly dependence on third source information.…”

Section: Introductionmentioning

confidence: 99%

A computational analysis of short sentences based on ensemble similarity model

Alhadi¹,

Deraman²,

Jalil³

et al. 2019

IJECE

View full text Add to dashboard Cite

<p>The rapid development of Internet along with the wide use of social media applications produce huge volume of unstructured data in short text form such as tweets, text snippets and instant messages. This form of data rarely contains repeated word. It presents challenge in sentences similarity analysis as the standard text similarity models merely rely on the number of word occurrence, often resulting unreliable similarity value. Besides, the use of abbreviation, acronyms, slang, smiley, jargon, symbol or non-standard short form also contributes to the difficulty in similarity analysis. Thus, an extended ensemble similarity model approach is proposed. An experimental study has been conducted using datasets of English short sentences. The findings are very encouraging in improving the similarity value for short sentences.</p>

show abstract

Matching, Re-Ranking and Scoring: Learning Textual Similarity by Incorporating Dependency Graph Alignment and Coverage Features

Kohail

Biemann

2018

Computational Linguistics and Intelligent Text Processing

View full text Add to dashboard Cite

A Fast Approach for Semantic Similar Short Texts Retrieval

Cited by 5 publications

References 19 publications

Bew: Towards Answering Business-Entity-Related Web Questions

Bew: Towards Answering Business-Entity-Related Web Questions

A computational analysis of short sentences based on ensemble similarity model

Matching, Re-Ranking and Scoring: Learning Textual Similarity by Incorporating Dependency Graph Alignment and Coverage Features

Contact Info

Product

Resources

About