arXiv subjects

Sahisnu Mazumder

Publications and source records attributed to Sahisnu Mazumder.

21 records · Page 2Linked to original sources

arXiv2018

Disentangling Aspect and Opinion Words in Target-based Sentiment Analysis using Lifelong Learning

Given a target name, which can be a product aspect or entity, identifying its aspect words and opinion words in a given corpus is a fine-grained task in target-based sentiment analysis (TSA). This task is challenging, especially when we have no labeled data and we want to perform it for any given domain. To address it, we propose a general two-stage approach. Stage one extracts/groups the target-related words (call t-words) for a given target. This is relatively easy as we can apply an existing semantics-based learning technique. Stage two separates the aspect and opinion words from the grouped t-words, which is challenging because we often do not have enough word-level aspect and opinion labels. In this work, we formulate this problem in a PU learning setting and incorporate the idea of lifelong learning to solve it. Experimental results show the effectiveness of our approach.

cs.CL↗

arXiv2017

Context-aware Path Ranking for Knowledge Base Completion

Knowledge base (KB) completion aims to infer missing facts from existing ones in a KB. Among various approaches, path ranking (PR) algorithms have received increasing attention in recent years. PR algorithms enumerate paths between entity pairs in a KB and use those paths as features to train a model for missing fact prediction. Due to their good performances and high model interpretability, several methods have been proposed. However, most existing methods suffer from scalability (high RAM consumption) and feature explosion (trains on an exponentially large number of features) problems. This paper proposes a Context-aware Path Ranking (C-PR) algorithm to solve these problems by introducing a selective path exploration strategy. C-PR learns global semantics of entities in the KB using word embedding and leverages the knowledge of entity semantics to enumerate contextually relevant paths using bidirectional random walk. Experimental results on three large KBs show that the path features (fewer in number) discovered by C-PR not only improve predictive performance but also are more interpretable than existing baselines.

cs.CL↗

arXiv2016

LANet: An Enriched Knowledgebase for Location-aware Activity Recommendation System

Accumulation of large amount of location-specific reviews on web due to escalating popularity of Location-based Social Networking platforms like Yelp, Foursquare, Brightkite etc. in recent years, has created the opportunity to discover location-specific activities and develop myriads of location-aware activity recommendation applications. The performance and popularity of such recommendation applications greatly depend on the richness and accuracy of the back-end knowledgebase, which intern is regulated by information relevancy and redundancy issues. Existing work on activity discovery have not made any attempt to ensure relevancy and non-redundancy of discovered knowledge (i.e., location-specific activities). Moreover, majority of these work have utilized body-worn sensors, images or human GPS traces and discovered generalized activities that do not convey any location-specific knowledge. In this thesis, we address the mentioned issues with serious concern and propose an effective solution to discover Location-specific Activity Network, in short LANet from location-aware reviews. The information network LANet serves as an accurate, enriched and unified knowledgebase of a Location-aware Activity Recommendation System. While building LANet, we also introduce novel ideas like, activity-based location similarity detection and measuring uniqueness, generality/speciality of an activity at a particular location to enrich the said knowledge base to a great extent. Experimental results show the information richness and accuracy of the proposed knowledge base which is comparable to human perception and accounts for our success in achieving the desired solution.

cs.SI↗