Computational Analysis and Deep Learning for Medical Care. Группа авторов
Чтение книги онлайн.
Читать онлайн книгу Computational Analysis and Deep Learning for Medical Care - Группа авторов страница 23
Название: Computational Analysis and Deep Learning for Medical Care
Автор: Группа авторов
Издательство: John Wiley & Sons Limited
Жанр: Программы
isbn: 9781119785736
isbn:
The primary goal of the search engines is to find the more appropriate or proximity document from the search engine. The basic idea of a search engine is a rudimentary search that will find the documents and rank the documents which are found in the search criteria. The suggestion engine uses a popularity score to determine which queries to suggest. With the exponential growth of the internet, searching for resources on the web (like, data, files, articles, etc.) is very common these days. Search engine efficiency is becoming a key factor in the web search. The effective way of improving the efficiency of the search engine is by using the keyword query suggestion [16]. Keyword suggestion is the most fundamental feature of the search engine. Users normally submit short queries to the web search engine, and short queries are mostly ambiguous. The major problem of the current web search engine is that search queries, as they are short. Users try different queries to retrieve the relevant information because the user may have little knowledge about the information of searching. The provided list of keywords by the web search engine may not be a good description of the information needs of the user [18-20].
The primary goals of the search engines are
1 Effectiveness (quality)
2 Efficiency (speed)
The growth and the significance of any search engines depend on these parameters only. Once the quality and speed of the search engines improve, the system performance will improve. Current research is toward the development of these performance measures. The success of any search engine is in a large part determined by the fact whether a user can find a good answer for his search query or not. That is why the most important aim of every search engine is to continuously improve its search performance. A lot of different techniques, architectures, algorithms, and models were invented and implemented to provide accurate search results that users consider as relevant and interesting. The basic model of the search engine has shown in Figure 2.1.
To swamp this problem, many search engines have implemented the query suggestion method. Also, known as keyword suggestion. The effective method for keyword suggestions is based on data from the query log [1]. This log is maintained by the search engines from the previous queries. These logs maintain lots of data with page ranks and the server address [15, 24, 25]. Location is not maintaining in many of the log databases. Need to implement the location of the user in some specific queries. In the scenario of a user that is searching for food in the afternoon, we need to suggest a hotel that is nearby his location, if the same query is asked in the morning session, we need to suggest a good hotel that serves breakfast. The spatial location of the user is critical in this case. The query suggestion is along with the location is important, to support more accurate results [17]. The main goal of the spatial keyword is to suggest more effortlessly to find appropriate results that will placate all the situations concerning the circumstances of a search. Searching motivated to develop methods to recover spatial objects.
Figure 2.1 General architecture of a search engine.
The main aim of the Artificial intelligence (AI) in the query suggestion is to automate the query suggestions based on the user circumstances. AI agents will learn the things based on the previous user preferences and locations; based on this, it will automate a query to the search engine and it recommends more accurate results to the user. It will help the user like a guide in specific applications based on his preferences. The AI agent learns the things from the user’s data and frames the appropriate query to get accurate results.
The advent of mobile devices and access to the internet in these devices makes us search location-based queries. Enhanced growth in the usage of the mobility devices has increased as shown in Figure 2.2. In the early age of computing, searches engines worked only on the keyword query-based searches. But, with the respect of these locations changes us to prompt for the location-aware keyword query is more significant these days. The mobile apps for transportation like Uber, Ola, and food delivery apps like Swiggy and Zomato are concentrating on the location of the user. Most of these apps are using the location-aware keywords only.
Figure 2.2 The increased mobile users.
AI-powered search engines learn from user behavior in real-time and reduce the gap between human and computer languages. The evolution of man and AI is taking place together. Making the user’s work still easier, the search strategies have been improved from taking the text as input to even voice and images. It is a proven fact that organizations with a documented personalized strategy have exceeded their revenue goals. AI perspective can serve the key purpose of adding convenience to the user. AI perspective also helps us to build upon the user’s reviews as well. For example, a user writes a review saying a movie or a journal lags in a particular aspect. The website or search engine should be able to make the movies or journals showing up as per the user’s choice.
2.2 Related Work
Informational retrieval is started from the creation of the HTML pages in the year 1945, organizing the documents a proper order and retrieve them. After the invention of the internet and the servers, it is a prominent problem for everyone. At the early age of the computers and internet, it is not a tough problem. In the early 90s, the exponential growth of the servers and connected devices to the internet has started the actual problem of referencing each other. The introduction of the HTTP and FTP creates more number of pages and stored on the internet. Along with the rapid developments of these technologies, information retrieval has become the crucial one. Gerard Salton [27] has started in this direction of the research, and we call him as the father of search engines. Their team started the work Automatic Retriever of Text, vector space model, Inverse Document Frequency (IDF), and Term Frequency (TF). The first search engine created was Archie, created in 1990 by Alan Emtage [28]. Search engine functionality depends on different aspects. Majorly, it consists of three main parts:
1 Spiders
2 Index
3 Search interface and relevancy software
The indexed data will be used for retrieving the information. It consists of different approaches, like ranking and string matching. The proximity measures are used in the literature for retrieving the documents but the retrieval in these cases has become a crucial aspect. The distance is like matching the keyword only. We need to consider the conditional probability of the keyword weather how accurately it is mapping to the given document. It will increase the proximity of the documents. Spatial objects also include spatial data along with longitude and latitude of the location. Many functionalities of a spatial database are helpful in distinct ways in specific contexts.
According to a study, nearly 75% of people are not satisfied with the information which they are suggested, that is not specific to them i.e., that is not personalized and not localized. The importance of AI perspective in this context can be explained with the help of an example. Let us say, a college student bought a laptop from an e-commerce website. Later, he wants to buy the add-ons for a laptop like a laptop cover, etc. Right now, it would be simply awesome if our website or search engine ranks up the suggestions of СКАЧАТЬ