Dialogue Intent Classification with Long Short-Term Memory Networks

Lian, Meng; Huang, Minlie

doi:10.1007/978-3-319-73618-1_4

Cited by 26 publications

(18 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our model we have combined the pre-processing techniques with a topic modelling Approach "Bag of words" to increase the accuracy of our system to some more extent. L. Meng [11] made use of LSTM along with CNN leading to a Deep CNN model which definitely had advantage over our model leading to 99% accu-racy, our model has 77.6% accuracy but for this little advancement, the overall complexity of the system became highly complex. Moreover, it was restricted to two topics(model proposed by Meng) whereas our model covered more than 10 topics leading to the larger dataset over the former.…”

Section: Comparative Analysismentioning

confidence: 99%

See 1 more Smart Citation

Text-Based Intent Analysis using Deep Learning

Desai¹,

Wadhwa²,

Anurag³

et al. 2020

IJISRT

View full text Add to dashboard Cite

Intent classification focuses on analyzing input to gauge the users’ needs and opinions. With this proposed system, we utilise a neural network classifier that analyzes text strings that are fed to it as input. Based on the analyzed intent, the system responds to the user by acting upon the commands issued to it. The areas covered by the system include weather, ratings, and booking restaurants. Firstly the intent of the text will be analysed and then it will be classified based on the topic it belongs to. Due to expressiveness of intent by users intent , for instance ”need a rental space”, ”places to visit in chennai” intent analysis plays a major role in ecommerce, real estate and online marketing industry. We aim to maximize the accuracy and minimize the errors and the time complextity of the proposed system

show abstract

Section: Comparative Analysismentioning

confidence: 99%

“…Lian Meng [11] used the hierarchical LSTM with an augmented memory module to utilize contextual information for dialog classification. The advantages over using an LSTM over an RNN for classifying long distance information.…”

Section: Robertomentioning

confidence: 99%

Text-Based Intent Analysis using Deep Learning

Desai¹,

Wadhwa²,

Anurag³

et al. 2020

IJISRT

View full text Add to dashboard Cite

show abstract

“…There is a wide variety of applications where activity, plan, and goal recognition algorithms have been applied and proven to be useful. Some examples are smart homes (Skocir et al, 2016), personal agent assistants (Oh et al, 2014), human-robot interaction (Kelley et al, 2010), video surveillance (Poppe, 2010), video games (Ha et al, 2014), natural language understanding (Meng and Huang, 2018), assistive care for the elderly (Bouchard et al, 2007), software help systems (Horvitz et al, 1998), computer network security (Rahmat et al, 2018), FIGURE 2 | Examples of scenarios of different characteristics, (A) is a grid navigation scenario (agnostic, no intervention, fully observable, deterministic, discrete, single agent), (B) is a poker game (adversarial, direct communication, partially observable, stochastic, discrete, multiagent), (C) is a platform video game (agnostic, no intervention, partially observable, deterministic, continuous, single agent), (D) is a human-robot collaboration scenario (intended, online intervention, partially observable, stochastic, continuous, single agent). decision support systems (Sengupta et al, 2017), and orthotics (Rebelo et al, 2013), among others.…”

Section: System Classificationmentioning

confidence: 99%

Section: The Problem Of Activity Plan and Goal Recognitionmentioning

confidence: 99%

Activity, Plan, and Goal Recognition: A Review

Van-Horenbeke

Peer

2021

Front. Robot. AI

View full text Add to dashboard Cite

Recognizing the actions, plans, and goals of a person in an unconstrained environment is a key feature that future robotic systems will need in order to achieve a natural human-machine interaction. Indeed, we humans are constantly understanding and predicting the actions and goals of others, which allows us to interact in intuitive and safe ways. While action and plan recognition are tasks that humans perform naturally and with little effort, they are still an unresolved problem from the point of view of artificial intelligence. The immense variety of possible actions and plans that may be encountered in an unconstrained environment makes current approaches be far from human-like performance. In addition, while very different types of algorithms have been proposed to tackle the problem of activity, plan, and goal (intention) recognition, these tend to focus in only one part of the problem (e.g., action recognition), and techniques that address the problem as a whole have been not so thoroughly explored. This review is meant to provide a general view of the problem of activity, plan, and goal recognition as a whole. It presents a description of the problem, both from the human perspective and from the computational perspective, and proposes a classification of the main types of approaches that have been proposed to address it (logic-based, classical machine learning, deep learning, and brain-inspired), together with a description and comparison of the classes. This general view of the problem can help on the identification of research gaps, and may also provide inspiration for the development of new approaches that address the problem in a unified way.

show abstract

“…It is noticeable that among all four corpora mentioned, there are no works that are applicable to the e-commerce setting. As of the time of writing there is only one e-commerce related work on DA classification by Meng and Huang (2017), which used a proprietary Chinese conversational dataset from a Chinese e-commerce service, however the dataset is not publicly available and details regarding its data collection were not specified. The lack of data for e-commerce dialogs motivated the building of the corpus for this work.…”

Section: Related Workmentioning

confidence: 99%

Annotation Process for the Dialog Act Classification of a Taglish E-commerce Q\&A Corpus

Rivera¹,

Pensica²,

Valenzuela³

et al. 2019

Proceedings of the Second Workshop on Economics and Natural Language Processing

View full text Add to dashboard Cite

With conversational agents or chatbots making up in quantity of replies rather than quality, the need to identify user intent has become a main concern to improve these agents. Dialog act (DA) classification tackles this concern, and while existing studies have already addressed DA classification in general contexts, no training corpora in the context of e-commerce is available to the public. This research addressed the said insufficiency by building a text-based corpus of 7,265 posts from the question and answer section of products on Lazada Philippines. The SWBD-DAMSL tagset for DA classification was modified to 28 tags fitting the categories applicable to e-commerce conversations. The posts were annotated manually by three (3) human annotators and preprocessing techniques decreased the vocabulary size from 6,340 to 1,134. After analysis, the corpus was composed dominantly of single-label posts, with 34% of the corpus having multiple intent tags. The annotated corpus allowed insights toward the structure of posts created with single to multiple intents.

show abstract

Dialogue Intent Classification with Long Short-Term Memory Networks

Cited by 26 publications

References 9 publications

Text-Based Intent Analysis using Deep Learning

Text-Based Intent Analysis using Deep Learning

Activity, Plan, and Goal Recognition: A Review

Annotation Process for the Dialog Act Classification of a Taglish E-commerce Q\&A Corpus

Contact Info

Product

Resources

About