Sentiment Analysis (detecting document's polarity, subjectivity and emotional states) is a difficult problem and several times I bumped into unexpected and interesting results. One of the strangest things that I found is that despite the fact that neutral class can improve under specific conditions the classification accuracy, it is often ignored by most researchers.

As it turns out, other techniques including website path analysis, text analysis of customer feedback, sentiment analysis of social media, and graph analysis --all distinctly different analytics techniques with each delivering insights complementing the others--revealed a fuller picture: people weren't complaining about price, preferring the cheaper item, or any of the things that the retailer expected. Instead, customers were complaining about how hard it was to find designer jeans on the website. It was a website navigation issue. And the issue was invisible until the retailer made sense of analytics from a variety of sources.

The veteran provider of document capture and OCR software has a new product suite aimed at enabling business applications to read and understand natural language, and thus help their users make better-informed decisions. Is the time right for a resurgence of interest in text analytics?

This blog was originally featured on blog.aylien.com, a Text Analysis blog with tutorials, Data Visualisations and industry discussions. Our founder, Parsa Ghaffari, gave a talk recently on Natural Language Processing and Sentiment Analysis at the Science Gallery in Dublin. As part of the talk, he put together a nice little example of how you can transform your Google Spreadsheet into a powerful Text Analysis and Data Mining tool. In this case, he took a simple example of analyzing restaurant reviews from a popular review site but the same could be done for hotels, products, service offerings and so on. He wanted to show how easy it can be for data geeks and even the less technical marketers among us, to start analyzing text and gathering business insight from the reams of textual data online today.

Some people seem to get along with everyone. Most people have an innate ability to read other people and quickly adjust their own style to match. We all do it to some degree, but some of us do it better than others. Consciously and unconsciously, we read other people by looking for cues – we evaluate facial expressions, vocal tone, body language, posture and gestures to assess how similar theirs are to our own. Sometimes we adjust accordingly and sometimes we don't, but the real purpose of reading other people is to reduce the ambiguity of our social interactions Great communicators exploit this phenomenon to build rapport with virtually anyone.

Data scientists must always remember that data sets are not objective - they are selected, collected, filtered, structured and analyzed by human design. Naked and hidden biases in selecting, collecting, structuring and analyzing data present serious risks. For example, a recent Wall Street Journal article entitled "Tweets Provide New Way to Gauge TV Audiences" provides evidence of a disconnect between mainstream viewers and folks who use Twitter. The chart above shows the disconnect between the most popular and most tweeted shows - the most tweeted show is not a top ten show. While Twitter data can be useful for detecting trends and sentiments for certain areas (e.g., disease surveillance, natural disaster surveillance, product sentiments, financial trading, politics) in limited circumstances using scientific methods, it can also mislead and present a false view of reality.

Natural language processing (NLP) involves machine learning, artificial intelligence, algorithms and linguistics related to interactions between computers and human languages. One important goal of NLP is to design and build software that will understand and analyze human languages to simplify and optimize human - computer communication. NLP algorithms are usually based on probability theory and machine learning grounded in statistical inference -- to automatically learn rules through analysis of real-world usage. It includes word and sentence tokenization, text classification and sentiment analysis, spelling correction, information extraction, parsing, meaning extraction, question answering and requires both syntactic and semantic analysis at various levels. NLP applications today involve spelling and grammar correction in word processors, machine translation, sentiment analysis and email spam detection.

MeaningCloud is an easy, powerful and affordable way to extract the meaning of any kind of unstructured content, from social conversations to internal files. It leverages a mix of natural language processing and machine learning technologies to provide APIs for information extraction, text classification and clustering, sentiment analysis, POS tagging… and other high-level functions like user profiling. These APIs can be totally customized to your domain using graphical tools and without the need to code, providing an unparalleled accuracy. And it features an add-in for Excel, so that you can do text analytics in your spreadsheet and a generous Free plan..

For a more technical explanation, this and this article can be read. Here you can find a good explanation as well as a list of the mostly used Kernel functions.

Meta-analysis is a principled statistical approach for summarizing quantitative information reported across studies within a research domain of interest. Although the results of meta-analyses can be highly informative,the process of collecting and coding the data for a meta analysis is often a labor-intensive effort fraught with the potential for human error and idiosyncrasy. This is due to the fact that researchers typically spend weeks poring over published journal articles, technical reports, book chapters and other materials in order to retrieve key data elements that are then manually coded for subsequent analyses (e.g., descriptive statistics, effect sizes, reliability estimates, demographics, and study conditions).In this paper, we propose a machine learning based system developed to support automated extraction of data pertinent to STEM education meta-analyses, including educational and human resource initiatives aimed at improving achievement, literacy and interest in the fields of science, technology, engineering, and mathematics.