Personal Attribute Prediction from Conversations

TL;DR

This paper introduces DSCGN, a noise-robust framework that predicts personal attributes from conversations without labeled data, leveraging external resources and unlabeled utterances to improve PKB enrichment.

Contribution

It proposes a novel, label-free method using distant and contextual supervision to effectively utilize personal knowledge embedded in conversations and external data.

Findings

Outperforms 12 baselines in nDCG and MRR metrics

Effective in predicting difficult personal attributes

Utilizes unlabeled data and external resources efficiently

Abstract

Personal knowledge bases (PKBs) are critical to many applications, such as Web-based chatbots and personalized recommendation. Conversations containing rich personal knowledge can be regarded as a main source to populate the PKB. Given a user, a user attribute, and user utterances from a conversational system, we aim to predict the personal attribute value for the user, which is helpful for the enrichment of PKBs. However, there are three issues existing in previous studies: (1) manually labeled utterances are required for model training; (2) personal attribute knowledge embedded in both utterances and external resources is underutilized; (3) the performance on predicting some difficult personal attributes is unsatisfactory. In this paper, we propose a framework DSCGN based on the pre-trained language model with a noise-robust loss function to predict personal attributes from conversations without requiring any labeled utterances. We yield two categories of supervision, i.e., document-level supervision via a distant supervision strategy and contextualized word-level supervision via a label guessing method, by mining the personal attribute knowledge embedded in both unlabeled utterances and external resources to fine-tune the language model. Extensive experiments over two real-world data sets (i.e., a profession data set and a hobby data set) show our framework obtains the best performance compared with all the twelve baselines in terms of nDCG and MRR.