|
| Sr. No. | Method used | Literature reference | Description | Justification in the context of research work |
|
| 1. | Markov model | [9–13] | It is a well-known approach for pattern recognition. It determines the next state from the current state based on the orders of the Markov chain | The main problem is lack of prediction accuracy with lower-order chain, while high complexity with the higher-order chain. However, this approach does not suit the current research context |
|
| 2. | Prediction by partial match | [15, 16, 33] | The PPM model uses a set of previous objects to predict the next item in a particular stream | It is a restricted version of Markov chain that provides prediction based on the only selected set of objects and selection of a right set of objects is a very challenging task, so this kind of vision is not also; it limits the result as it does not cover all the objects, thereby ruling it out of the scope of current work |
|
| 3. | Cost function | [14, 17] | Prediction of future requests has been made based upon certain factors like the popularity and lifetime of web objects | A very less popular approach for pattern determination as the cost functions vary from time to time, thereby reducing the contribution in making the right set of prediction. So this approach is also not suitable in the context of the proposed research |
|
| 4. | Data mining | [18] | It is also one of the most popular approaches in the modern era for pattern recognition of structured objects | The data mining approach consists of many techniques which are ideal for pattern generation task. But the proposed research is not working upon pattern generation task |
|
| 5. | Keyword based | [19, 20, 24, 25] | Prediction is made by retrieving confidential information present in the contents of web documents | To work upon only this category is not much beneficial since it does not deal with multiple user transactions |
|
| 6. | Integration of domain knowledge | [21–23] | It works by the integration of domain knowledge with other methods of prefetching; semantics are taken into account | It gives useful information based on semantics but increases prediction time as well as extra overhead |
|
| 7. | Implicit link analysis | [26, 29–32] | In the “implicit link analysis” technique, the importance of a web page is determined by the users who navigate the web page | It is a significantly less popular approach for pattern determination. Extra work is required to reorganize the structure of the website as per user navigation |
|
| 8. | Explicit link analysis | [26–28] | In the “explicit link analysis” technique, the importance has been given to the design that has been structured by the designer who makes any web page more important or less important | It gives useful information based on hyperlink structures of the web |
|
