Note: I don't know the techniques used by Microsoft Live/Bing (9/28/2007), but Google has a paper. To see the model, please check out (Hu and Liu, KDD-2004) and (Liu et al, WWW-2005) below, or the book above (better). Try search for a camera and click on reviews. You will see summarized user opinions on product features/aspects in a bar chart.
Opinion Parser: the sentiment analysis system used in the startup company, OpinionEQ.
Try It: If you give me a text file, I can run it for you. Each line of the file should be in the following format representing a sentence (the first number is the sentence id): 1020/CD It//PRP gave/give/VBD good//JJ results/result/NNS .//.
The system analyzes sentiments/opinions and emotions, extracts entities, topics and their aspects/features/attributes, and handles sentiments in comparative sentences. The system is automatic.
< Another 5 keynote and invited talks to be given later this year >
Keynote talk: Sentiment Analysis: Past, Present and Future. Conference on Behavioural Models & Sentiment Analysis Applied to Finance. London, UK, June 18-19 2014.
Invited talk: Aspect-based Sentiment Analysis. Workshop on Sentiment Classification and Opinion Mining Using News Wires and Micro-blogs, London, UK, June 20, 2014
Keynote talk: From Sentiment Analysis to Continous Learning (topic disocvery), Summer Research Insitute, EPFL - Lausanne, Switzerland, June 2 2014.
Several invited talks on sentiment analysis, machine learning and NLP were given in May 20, 26 and 27, 2014 in Jiangxi University of Finance and Economicis, Renmin University, Tsinghua University, and Beijing Univrsity of Posts and Telecommunications, China.
Keynote talk: Sentiment Analysis and Natural Language Processing. The 15th International Conference on Intelligent Text Processing and Computational Linguistics (CICling-2014), April 6-12, 2014, Kathmandu, Nepal.
Keynote talk: Combining Knowledge and Probabilistic Inference for Topic Extraction. The second international conference on Big Data Analytics, Dec. 16-18, India.
Keynote talk: Detection of Fake or Deceptive Opinions. International Winter School on Machine Learning and Text Analytics, Dec. 19, 2013, India.
Keynote talk: Sentiment centric analysis of social media content, The 26th Conference on Computational Linguistics and Speech Processing, Oct 4, 2013, Taiwan.
Invited lecture: Sentiment analysis and opinion mining. 12th Estonian Summer School on Computer and Systems Science, Aug 18 - 22, 2013, Voore Guesthouse, Estonia.
Invited talk: Sentiment Analysis and Modeling of Discussions. Peking University, Beijing, May 16, 2013.
Invited talk: Sentiment Analysis and Social Media Modeling. Institute of Software, Chinese Academy of Sciences, May 13, 2013, Beijing.
Invited talk: Detecting Deceptive and Fake Opinions on the Web. Institute of Software, Chinese Academy of Sciences, May 14, 2013, Beijing.
Invited talk: Sentiment Analysis and Modeling of Discussions. Midwest Speech and Language Days, May 2-3, 2013, Chicago.
Keynote talk: From Opinion Mining to Modeling of Comments and Debates.
Conference for Big Data Systems, Applications and Privacy, March 10-11, 2013, Abu Dhabi.
Invited talk: Modeling of Comments and Debates. IKDD workshop (Inaugural Workshop of India KDD Chapter), Feb 15, 2013, Mysore, India.
Invited talk: Sentiment Analysis and Social Media Modeling. TCS Symposium, Feb 12, 2013, New Delhi, India.
Invited talk: Sentiment Analysis and Social Media Modeling. The Marketing Department, Kellogg School of Management, Northwestern University, January 23, 2013.
Invited talk: Analysis and Modeling of Opinions and Sentiments. Toyota Technological Institute at Chicago (TTI). Dec 4, 2012, Chicago, USA.
Invited talk: Modeling Opinions and Beyond in Social Media. KDD Summer School (Aug 10, 2012) at the KDD conference (Aug 12-16, 2012), Beijing, China. (This is an 2-hour lecture. I will talk about how to model opinions, comments, debates, etc. in social media).
Invited talk: Analysis and Modeling of Sentiments and Opinions. IBM Almedan Research Center, June 21, 2012.
This work is in the area of sentiment analysis and opinion mining from social media, e.g., reviews, forum discussions, and blogs. In our KDD-2004 paper, we proposed the Feature-Based Opinion Mining model,
which is now also called Aspect-Based Opinion Mining
(as the term feature here can confuse with the term feature used in
machine learning). The output of such opinion mining is a
feature-based opinion summary
or aspect-based opinion summary. The commonly known sentiment classification is a sub-task. Our current work is in two main areas,
which reflect two kinds of opinions (or evaluations)
Mining regular (or direct) opinions. Ex: (1). This camera is great. (2). After taking the drug, I got stomach pain.
Mining comparative opinions. Ex: Coke tastes better than Pepsi.
2. Sentiment Analysis or Mining of Regular Opinions
In this research, we aim to mine and to summarize online opinions in reviews,
tweets, blogs, forum discussions, etc. Specifically, we mine features or aspects of entities (e.g., products) or topics on which people have expressed their opinions and determine whether the opinions are positive or negative. For opinion summarization, we advocate the quantitative aspect and the target of opinions because 50% of the people say something is bad is not the same as 5% say it is bad.
Amazon Product Review Data (more than 5.8 million reviews) used in (Jindal and Liu, WWW-2007, WSDM-2008; Lim et al, CIKM-2010; Jindal, Liu and Lim, CIKM-2010; Mukherjee et al. WWW-2011; Mukherjee, Liu and Glance, WWW-2012) for opinion spam (fake review) detection. You can also use it for sentiment analysis. It has information about reviewers, review texts, ratings,
product info, etc. Due to the large file size, you may need to use Download Accelerator Plus (DAP) to download. If you use this data, please cite (Jindal and Liu, WSDM-2008).
Pros and cons dataset used in (Ganapathibhotla and Liu, Coling-2008) for determining context (aspect) dependent sentiment words, which are then applied to sentiment analysis of comparative sentiences (comparative sentence dataset). The same form of Pros and Cons data was also used in (Liu, Hu and Cheng, WWW-2005).
Zhiyuan Chen, Arjun Mukherjee, Bing Liu, Meichun Hsu, Malu Castellanos, and Riddhiman Ghosh. Exploiting Domain Knowledge in Aspect Extraction. Proceedings of Conference on Empirical Methods in Natural Language Processing (EMNLP-2013), October 18-21, 2013, Seattle, USA.
Zhiyuan Chen, Bing Liu, Meichun Hsu, Malu Castellanos, and Riddhiman Ghosh. Identifying Intention Posts in Discussion Forums. Proceedings of The 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT-2013), June 9-15, 2013, Atlanta, USA.
Lei Zhang and Bing Liu. "Extracting Resource Terms for Sentiment Analysis," Proceedings of the 5th International Joint Conference on Natural Language Processing (IJCNLP-2011), November 8-13, 2011, Chiang Mai, Thailand.
Minqing Hu and Bing Liu. "Mining and summarizing customer reviews."Proceedings of the ACM SIGKDD International Conference on
Knowledge Discovery & Data Mining (KDD-2004, full paper), Seattle,
Washington, USA, Aug 22-25, 2004.
Arjun Mukherjee, Abhinav Kumar, Bing Liu, Junhui Wang, Meichun Hsu, Malu Castellanos, and Riddhiman Ghosh. Spotting Opinion Spammers using Behavioral Footprints. Proceedings of SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-2013), August 11-14 2013 in Chicago, USA.
Arjun Mukherjee, Vivek Venkataraman, Bing Liu, and Natalie Glance. What Yelp Fake Review Filter Might Be Doing. Proceedings of The International AAAI Conference on Weblogs and Social Media (ICWSM-2013), July 8-10, 2013, Boston, USA.
Nitin Jindal and Bing Liu. "Opinion Spam and Analysis."Proceedings of First ACM International Conference on Web Search and Data Mining (WSDM-2008), Feb 11-12, 2008, Stanford University, Stanford, California, USA.