Zhang, Wei (Tsinghua University and Tsinghua National Laboratory for Information Science and Technology) | Wang, Jianyong (Tsinghua University and Tsinghua National Laboratory for Information Science and Technology)
User-item connected documents, such as customer reviews for specific items in online shopping website and user tips in location-based social networks, have become more and more prevalent recently. Inferring the topic distributions of user-item connected documents is beneficial for many applications, including document classification and summarization of users and items. While many different topic models have been proposed for modeling multiple text, most of them cannot account for the dual role of user-item connected documents (each document is related to one user and one item simultaneously) in topic distribution generation process. In this paper, we propose a novel probabilistic topic model called Prior-based Dual Additive Latent Dirichlet Allocation (PDA-LDA). It addresses the dual role of each document by associating its Dirichlet prior for topic distribution with user and item topic factors, which leads to a document-level asymmetric Dirichlet prior. In the experiments, we evaluate PDA-LDA on several real datasets and the results demonstrate that our model is effective in comparison to several other models, including held-out perplexity on modeling text and document classification application.
Consumers now experience AI mostly through image recognition to help categorize digital photographs and speech recognition that helps power digital voice assistants such as Apple Inc's Siri or Amazon.com But at a press briefing in San Francisco two days before Ng's Landing.ai In many factories, workers look over parts coming off an assembly line for defects. Ng showed a video in which a worker instead put a circuit board beneath a digital camera connected to a computer and the computer identified a defect in the part. Ng said that while typical computer vision systems might require thousands of sample images to become "trained," Landing.ai's
Amazon's voice assistant Alexa can order an Uber, check your credit card balance and even arrange for a Domino's pizza to be delivered to your doorstep. Next up, it will be able to lock your door, thanks to an integration with August, a startup that makes Internet-connected door locks. Embedded in Amazon's voice-controlled Echo smart speakers, as well as its Fire TV video streaming device, Alexa answers questions and lets users order items like diapers from Amazon.com. Over the past year, a number of home automation and smart device companies have built their own skills (the Amazon name for Alexa's apps) for Alexa, letting users turn their lights on or off, or control their thermostats with the voice assistant. Launched in 2014, August's smart lock lets homeowners essentially connect their lock to their smartphone, turning their phones into virtual keys.
In a recent report, the National Retail Federation projected online sales in 2017 will grow three times faster than in-store sales. The report suggests 51% of Americans prefer to shop online rather than in stores--a figure that jumps to 67% for millennials and 56% for Gen Xers. With Amazon accounting for 43% of online sales in the US, only store sales will be disrupted more in the coming months. Additionally, eCommerce sales are also expected to reach $4 trillion by 2020 – making it 14.6% of total retail spending that year. So what does this mean for retailers?
Computers may not wear tennis shoes (yet), but thanks to developing artificial intelligence technologies, they're smarter than ever before. Along with those technologies has come a relatively new category of computer science called machine learning, or ML. Similar to statistics, ML involves computer systems that utilize algorithms to automatically learn about data, recognize patterns, and make decisions, all without outside intervention or explicit directions from human beings. In the real world, you can find it being used in smart assistants like Siri and the Amazon Echo, in online fraud detection services, in the facial recognition feature that identifies photos of you on Facebook, and more recently, in Tesla's self-driving car. ML is distinctive in the world of AI in that it can be used to process vast amounts of data quickly, making it a desirable tech skill among job applicants not only in the fields of computer science and engineering, but also marketing, health care, finance, social media, and beyond.