By super nervous
Scenario Description: Applying AI technologies such as computer vision, speech recognition, natural language processing and big data analysis to the tourism service industry can help tourists reduce unnecessary waste of time and money on the one hand, and improve the service efficiency and quality of scenic spots to achieve a win-win situation.
Keywords: crawler, computer vision, speech recognition, natural language processing, big data analysis, cloud computing
The May Day holiday is over, have you calculated where you spent your holiday time? Traffic jam? Line up for attractions? Or wait for a C-position to take a picture?
It is said that the four-day holiday gives tourists a Spring Festival travel experience. Train tickets and popular scenic spots are hard to get, and scalpers take the opportunity to make a lot of money:
Shanghai Disneyland was so popular during the May Day holiday,
Ticket prices have been doubled
Someone joked that he had finally understood Thanos after returning from the May Day holiday.
Data analytics help tourists avoid the crowd
In the face of this situation, the smart programmer thought of a way to help everyone tact to avoid the sea of people.
After analyzing the data obtained by the crawler, the programmer decided that Shanghai Disneyland was the most popular attraction on May Day this year, so he decided not to join in the fun.
After analyzing the data of some popular scenic spots in the city, get the popularity ranking and see if you contributed to the data
The rankings were based on an analysis of ticket sales in top destinations on Qunar. Specific methods are as follows:
By requesting https://piao.qunar.com/ticket/list.htm?keyword= Beijing, popular scenic spot in the Beijing area information, and then by analyzing BeautifulSoup to extract the needed information.
According to the climbing data, not only can the ticket sales of scenic spots be analyzed, but also the thermal map of scenic spots can be made, taking Beijing as an example:
Scenic spots such as the Palace Museum, Gong Wang Fu Palace and Beijing Zoo were the most popular during this year’s May Day holiday, according to the chart
The open platform of Baidu Map is used here. First of all, you need to register the developer information. At the bottom of the home page, there is a “Apply for secret key” button, click to create.
Since the application type is browser-side, you only need to assemble the data and replace the corresponding HTML code. You also need to replace the AK that you use to access your application.
Finally, the data are also used to compare the prices of popular scenic spots so as to make corresponding plans according to their own budgets.
Another programmer used Python to climb the data of scenic spots in all provinces and autonomous regions of China, and obtained the national scenic spot heat map during last Year’s National Day:
It can be clearly seen from the figure that the thermal values of Beijing-Tianjin-Hebei region, Yangtze River Delta, Pearl River Delta and Sichuan region rank the top four
Armed with this analysis, you can plan your next vacation so you don’t have to worry about getting stuck in traffic.
AI helps scenic spots improve efficiency
In addition to tourists, scenic spots and food shops are also plagued by huge crowds during long holidays.
During the May Day holiday, several scenic spots were blocked, and the Badaling Great Wall received 54,000 tourists on the first day of the holiday, generating nearly 200,000 tons of garbage. The expo, which has just opened, welcomed 327,000 visitors, and its spectacular picture can be imagined.
Tourists on the Great Wall experience a dilemma
An Internet celebrity lobster restaurant in Changsha issued a letter of apology, saying it could not receive so many customers, but on its list, it had a table of 7,172.
Faced with these problems, many companies are also actively using ARTIFICIAL intelligence to help scenic spots find solutions.
The earliest AI + tourism starts from a single scene, hoping to promote the intelligence of the tourism industry. For example, in many scenic spots we will encounter voice tour/translation, robot customer service, AR/VR digital tour and so on. However, these apps are all independent services, and there is still no fundamental improvement for tourists’ experience.
In recent years, some domestic companies have taken a new step and started to provide comprehensive operational analysis basis for scenic spots and local tourism departments.
Aibee’s first project “Face brushing into the Garden” has been put into use in Kanas, Qin Emperor’s Terracotta Warriors and Horses, Gubei Water Town and other scenic spots
For example, Aibee, founded by Dr. Lin Yuanqing, former director of Baidu’s Deep Learning lab, combines AI with offline tourism to provide a holistic tourism solution.
They integrate online and offline data. First, visitors can register and buy tickets by swiping their faces. After entering the park, they will track their progress. During the tour, the scenic spot can use recommendation, popup and other forms of information push, service to find people, so as to promote the second consumption of tourists inside and outside the scenic spot.
Aibee’s solution involves multi-mode AI technology including computer vision, speech recognition, natural language understanding, big data analysis, etc. The technology is not complicated, but the integration of these technologies and the overall solution will provide great help for the improvement of service efficiency and quality of scenic spots.
At present, Aibee has cooperated with 65 scenic spots such as Wudang Mountain, Kanas mountain and Huashan Mountain, and its facial recognition system only takes a few seconds, while the traditional way of checking tickets takes at least 30 seconds, improving the efficiency of entering the park by about 10 times.
BAT AI + tourism layout
In the face of the tourism market, BAT can only be willing to be an audience?
Tencent: Last August, Tencent and Chongqing jointly created “A Mobile Phone Tour wulong”, involving the Internet of Things, cloud computing, big data, artificial intelligence and other technologies. Through personalized circuit customization, expert recommendation, intelligent customization, intelligent navigation and other functions, It provides one-stop intelligent services for tourists in terms of tourism information acquisition, itinerary planning, product booking, travel notes sharing, and featured e-commerce purchase.
Ali: Last year, FlyZoo Hotel (Chinese name: FlyZoo Hotel) opened its first physical store in the future. It is said to be the first Hotel in the world that supports full-scene face-scanning accommodation, realizing intelligent AI services from check-in, room experience and check-out process.
A robot delivers food at the Philippine Hotel
Baidu: Oriented to the competent department of tourism and scenic spot, baidu also actively develop its search, portrait, public opinion, knowledge map and data and technical advantages, such as artificial intelligence lines to tourists before, in and after full coverage and comprehensive tourism big data layout, provide visualization platform (SaaS) and data interface (API) services, help promote accurate marketing ability of scenic spots, Optimize the efficiency of safety management, improve the quality of tourist services, and guide the layout of regional tourism.
Baidu’s AI + tourism solution architecture
So, before you go out on your next vacation, you might as well use the power of technology to do a heat analysis of the scenic spots on your travel list, avoid the crowds of people tactfully, and stride over mountains and seas calmly.
Hyperneuropedia
Multimodal Learning
A modal is a particular way in which a person receives information. Since multimedia data is often the transmission medium of multiple information (for example, text information, visual information and auditory information are often transmitted simultaneously in a video), multimodal learning has gradually developed into the main means of multimedia content analysis and understanding.
Multimodal learning mainly includes the following research directions:
Multimodal representation learning: it mainly studies how to numerically transform the semantic information contained in multiple modal data into real-valued vectors.
Modal mapping: The study of how to map information from data in one particular mode to another mode.
Alignment: It mainly studies how to identify the correspondence between components and elements in different modes.
Fusion: mainly studies how to integrate models and features of different modes.
Collaborative learning: it mainly studies how to transfer the knowledge learned in the information-rich mode to the information-poor mode, so that the learning of each mode can assist each other. Typical methods include multi-modal zero-sample learning and domain adaptation.
Historical article (click on image to read)
My October holiday trip was completely controlled by artificial intelligence
Take advantage of holidays to practice internal skills, teach you to smoothly switch programming languages
Behind the Beijing Expo: The Beauty of horticulture brought by technology
AI encyclopedia
The tutorial
The data set
The store
More and more
http://hyper.ai
