Context-Aware Intent Identification in Email Conversations

Wang, Wei; Hosseini, Saghar; Awadallah, Ahmed Hassan; Bennett, Paul; Quirk, Chris

doi:10.1145/3331184.3331260

Cited by 33 publications

(19 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dialog acts identification could be seen as determining interlocutors intentions within a conversation. Wang et al (2019) study the identification of intentions in emails in a workplace situation.…”

Section: Future Workmentioning

confidence: 99%

Building A Corporate Corpus For Threads Constitution

Tadjou¹,

Bourge²,

Marie³

et al. 2021

Proceedings of the Student Research Workshop Associated With RANLP 2021

View full text Add to dashboard Cite

In this paper we describe the process of building a corporate corpus that will be used as a reference for modelling and computing threads from conversations generated using communication and collaboration tools. The overall goal of the reconstruction of threads is to be able to provide value to the collorator in various use cases, such as higlighting the important parts of a running discussion, reviewing the upcoming commitments or deadlines, etc.Since, to our knowledge, there is no available corporate corpus for the French language which could allow us to address this problem of thread constitution, we present here a method for building such corpora including different aspects and steps which allowed the creation of a pipeline to pseudo-anonymise data. Such a pipeline is a response to the constraints induced by the General Data Protection Regulation GDPR 1 in Europe and the compliance to the secrecy of correspondence.

show abstract

Section: Future Workmentioning

confidence: 99%

Building A Corporate Corpus For Threads Constitution

Tadjou¹,

Bourge²,

Marie³

et al. 2021

Proceedings of the Student Research Workshop Associated With RANLP 2021

View full text Add to dashboard Cite

show abstract

“…(1) Classical machine learning model baseline: we use an n-gram (up to 3-g) TF-IDF featurizer and a linear Support Vector Machine (SVM) model to build a classical machine learning baseline without using any regularization. This has been shown to be a strong non-neural network-based baseline model 21 for email intent classification. efficient, which is why it is often chosen over the more advanced LSTM layer, which is usually recommended.…”

Section: Experiments Runsmentioning

confidence: 99%

Transfer learning meets sales engagement email classification: Evaluation, analysis, and strategies

Liu¹,

Dmitriev²,

Huang³

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

Enterprise email classification in the sales engagement platform is a challenge due to its evolving asynchronous conversational context during the sales process and differences across industries and organizations. This is further exacerbated by the limited amount of labeled emails due to security and privacy constraints. The leaderboard success of using pretrained language models (LMs) such as BERT and various transfer learning techniques promises a paradigm shift to natural language processing, yet the recipe for applying high performance transfer learning (HPTL) in practical applications remains unclear. This article investigates applying HPTL to sales engagement email classification through a series of experiments and analysis. The experiment datasets include two different organizations' emails. The contribution of this paper is 4-fold: (a) analysis and characterization of the email corpora from different organizations; (b) identification of the best combinations of pre-trained LMs under different modeling architectures; (c) study of the impact and trade-off of limited labeled data on the model accuracy and training time; and (d) characterization and study of the impact of different orgs' datasets on the model accuracy. Our results showed that a practical winning recipe that uses BERT-finetuning with as few as 500 labeled training examples can consistently outperform significantly with reasonable training time among all models evaluated. K E Y W O R D S cross-org transfer learning, domain shift, email intent classification, pre-trained language model, sales engagement, transfer learning 1 INTRODUCTION Sales are one of the oldest professions on earth. 1 Until very recently, a typical sales representative (sales rep) got a list of names (leads, or prospects) and manually went through the list one by one calling and emailing the prospects. The rise of Sales Engagement Platforms (SEPs) such as Outreach, SalesLoft, InsideSales, Groove, and Apollo has rapidly changed this state of affairs, leading to large improvements in rep performance. SEPs encode a company's sales process into a sequence of steps consisting of emails, phone calls, LinkedIn messages, and other tasks. Different sequences are usedfor different types of prospects, market segments, and so forth. SEP then ensures consistent execution of these sequences, completely automating some sales tasks (eg, auto-sending personalized emails and LinkedIn messages), while scheduling and reminding the rep when it is the right time to do the manual tasks (eg, phone call, custom manual email). As a result, every rep can simultaneously perform one-on-one personalized outreach to up to 10× more prospects than before.

show abstract

“…Text classification is a fundamental task in text mining with a wide spectrum of applications such as text geolocalization [4], sentiment analysis [25], and email intent detection [33]. Following the routine of supervised learning, one can build a text classifier from humanannotated training documents.…”

Section: Introductionmentioning

confidence: 99%

MotifClass: Weakly Supervised Text Classification with Higher-order Metadata Information

Zhang

Garg

et al. 2022

Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining

View full text Add to dashboard Cite

We study the problem of weakly supervised text classification, which aims to classify text documents into a set of pre-defined categories with category surface names only and without any annotated training document provided. Most existing classifiers leverage textual information in each document. However, in many domains, documents are accompanied by various types of metadata (e.g., authors, venue, and year of a research paper). These metadata and their combinations may serve as strong category indicators in addition to textual contents. In this paper, we explore the potential of using metadata to help weakly supervised text classification. To be specific, we model the relationships between documents and metadata via a heterogeneous information network. To effectively capture higher-order structures in the network, we use motifs to describe metadata combinations. We propose a novel framework, named MotifClass, which (1) selects category-indicative motif instances, (2) retrieves and generates pseudo-labeled training samples based on category names and indicative motif instances, and (3) trains a text classifier using the pseudo training data. Extensive experiments on real-world datasets demonstrate the superior performance of MotifClass to existing weakly supervised text classification approaches. Further analysis shows the benefit of considering higher-order metadata information in our framework.

show abstract

Context-Aware Intent Identification in Email Conversations

Cited by 33 publications

References 19 publications

Building A Corporate Corpus For Threads Constitution

Building A Corporate Corpus For Threads Constitution

Transfer learning meets sales engagement email classification: Evaluation, analysis, and strategies

MotifClass: Weakly Supervised Text Classification with Higher-order Metadata Information

Contact Info

Product

Resources

About