Yelp Utilises Neural Networks to Enhance User Experience and Business Value

Yelp, the popular business directory service, has recently unveiled a series of advancements in its platform, tapping into the power of artificial intelligence (AI). These enhancements aim to provide increased value to both consumers and businesses alike. By harnessing the capabilities of neural networks, Yelp is set to redefine its search, ad matching, photo classification, and the much-touted Yelp Waitlist features.

For years, Yelp has been at the forefront of AI integration, having previously invested in machine learning models, XGBoost, and more recently, large language models (LLMs). Neural networks, known for their hierarchical, multi-layered structure, are particularly adept at extracting complex features from raw data. This has proven beneficial for Yelp, given its vast troves of user-generated content. The scalability of neural networks has also allowed the company to amplify its training sets, resulting in the development of enhanced models tailored for Yelp’s specific requirements.

One of the primary announcements from Yelp is the broadening application of neural networks to elevate its ad matching, search results, and photo classification, among other features. For the everyday user, this translates to search results that better align with their intent. It also promises ads that are more pertinent, photos that aptly represent businesses, and more precise wait time estimates through Yelp Waitlist.

Over the past year, Yelp has been fine-tuning its ad systems to more accurately pair ads with the right consumers. This is based on the context of the user’s search query. Neural networks have played a pivotal role in this endeavour, enabling the processing of years’ worth of data. This marks a significant leap from previous capabilities, where only months of data could be processed.

Enhancing Photo Relevance for Businesses and Consumers

Local businesses can also stand to gain from these advancements. The recent update from Yelp allows businesses to display photos within their ads. These can either be self-selected or chosen by Yelp. Through the use of neural networks, the content within these photos is intelligently identified. The best photos are then selected based on various criteria to match consumer search intent. This has led to a notable uptick in clickthrough rates and a reduction in cost per click for advertisers.

The integration of AI tools, including traditional machine learning models, LLMs, and neural networks, has further refined Yelp’s consumer search experience. This sophisticated blend ensures users are presented with the most relevant businesses and photos that align with their search queries. Neural networks have been instrumental in training models using years of photo data, which aids in providing quality scores for each image. The holistic appeal of photos is also considered, ensuring consumers are presented with the best representation of businesses.

Ensuring Content Quality and Safety

Yelp’s commitment to neural networks is not just a recent endeavour. The company has been utilising them for nearly a decade, predominantly for image classification. This has enabled consumers to easily locate photos that pique their interest, be it the latest menu addition or a specific cuisine. Neural networks have also been employed to sift through Yelp review text. This helps in identifying the most talked-about foods and beverages, which in turn informs the Popular Dishes and Drinks features.

Trust and integrity are paramount for Yelp. In its bid to create a secure environment for its users, the company has employed neural networks to detect photos that might breach content guidelines. Such images are flagged for human moderators to review before they are made live on the platform.

Optimising Yelp Waitlist with Neural Networks

Yelp Waitlist, an integral feature of Yelp Guest Manager, has also seen significant enhancements. Many consumers rely on this feature to gauge a restaurant’s live wait time. Neural networks have been pivotal in boosting the accuracy of these wait time estimates, benefitting both diners and businesses.

In conclusion, Yelp’s rich repository of user-generated content and intricate internal data has positioned it perfectly to leverage cutting-edge technology. With neural networks, LLMs, and other AI models at its disposal, the company is poised to introduce innovative solutions and features that will further amplify its platform. The future certainly looks promising, with more updates anticipated in the near future.