Analytics over the Hows and the Whys

You have reached the homepage of the ICWSM 2018 Tutorial titled Analytics over the Hows and the Whys . This tutorial is jointly prepared by Deepak P, Dinesh Garg and Shirish Shevade. This tutorial is now over; we have recorded the tutorial in videos which are enclosed below.

Contents Youtube Video

Part 1: Introduction, Challenges and Tasks

1. Distinguishing CQA Retrieval/Analytics from Factoid QA and Information Retrieval
2. Typical character of CQA Questions and Answers
3. Holy Grail of Computational CQA
4. Current State of Computational CQA Retrieval/Analytics
5. Types of Data available in CQA
6. Typical Operation of a CQA portal
7. Case-based Reasoning and Textual CBR and how they are similar to CQA
8. Tasks in CQA Analytics [Retrieval, Expert Finding, Organization, Maintenance and Procurement of QA Data]
9. Lexical Chasm as a Core Challenge in CQA Analytics
10. Word Correlation Learning from Parallel Data as a model for transcending the Lexical Chasm

Part 2: CQA Retrieval Methods

1. Overview of Technical Building Blocks used in CQA Retrieval methods
2. Introduction to statistical machine translation modeling basics
3. Paper: Retrieval Models for QA Archives, SIGIR 2008
4. Introduction to Topic Modelling
5. Paper: Question Retrieval with High-Quality Answers in CQA, CIKM 2014
6. Paper: Learning to suggest questions in social media, KAIS 2014
7. Introduction to Word Embeddings and Document Embeddings
8. Introduction to Auto-Encoders
9. Paper: Learning Semantic Representation with Neural Networks for CQA Retrieval, KBS 2016
10. Paper: Question/Answer Matching for CQA Systems ... , AAAI 2015
11. Introduction to Matrix and Tensor Decomposition fundamentals
12. Paper: Latent Semantic Tensor Indexing for CQA, ACL 2013
13. Introduction to Locally Linear Embedding
14. Paper: Latent Space Embedding for Retrieval in QA Archives, EMNLP 2017

Part 3: Other Tasks in CQA Analytics

1. QA Procurement from forum data
2. Generative Models for Problem solution pairs (from Paper: Two-part segmentation of text documents, CIKM 2012)
3. Paper: Unsupervised Solution Post Identification from Discussion Forums, ACL 2014
4. Question Quality Prediction
5. Paper: Great Question! Question Quality in Community Q&A, ICWSM 2014
6. Answer Quality Prediction
7. Paper: Discovering high quality answers in community question answering archives using a hierarchy of classifiers, Information Sciences 2014
8. Question Hardness Prediction
9. Paper: Question Difficulty Evaluation by Knowledge Gap Analysis, ASONAM 2014
10. Paper: MixKMeans: Clustering QA Archives, EMNLP 2016
11. Expert Finding in Social QA Sites
12. Introduction to the HITS Algorithm
13. Paper: Discovering Authorities in QA Communities using Link Analysis, CIKM 2007
14. Introduction to PageRank Algorithm
15. Paper: An empirical study of topic-sensitive probabilistic model for expert finding in question answer communities, Knowledge-based Systems 2014
16. Query Suggestions for QA
17. Paper: Query Suggestions for Textual Problem Solution Repositories, ECIR 2013

Part 4: Multi-view Analytics

1. QA as 2-view data
2. Examples of Multi-view Data
3. Examples of Multi-view processing in the mind
4. Core Tasks in Multi-view Learning
5. Framework of the challenge in multi-view learning
6. Learning a single latent space
7. Learning view-specific private latent spaces
8. Deep Learning for Multi-view Analytics Overview

Part 5: Potential Research Directions

Please write to for any queries or to offer any comments.