Based on the comprehensive literature survey, we have reached the point where we can safely say that children are still struggling from the very basic problems of spelling and typing mistakes, limited vocabulary, and making sense of results. Solutions exist here and there but each of them is far from best and leaves a lot of room for improvement. In the previous chapters we discussed the features and advantages of faceted navigation and it is evident that faceted search has certain features and qualities that it can prove helpful in solving the long lasting problems of children. It is just the matter of how to use it in a proper way. Now, we will present our proposed interface and its various elements and we will explain how these elements cater to the various problems of the children they encounter throughout their searching process.

3.1 Motivating Scenario

One of the major activities for children over the Internet is searching the web for the information related to their school work e.g. quizzes, exams, books, general knowledge, etc. He/she use different resources mainly search engines to find related information. Now let’s think of a scenario considering steps a child goes through to search for the required information. This scenario is described below.

Asjad studies in the 5th grade. In his science class the topic of today was space and teacher told the class about asteroids. At the end of the class teacher gave the class an assignment to go and search on the internet for more information about asteroids and come with a short essay in the next class. Asjad connects to the Internet after reaching home and opens his favorite search engine which is Google for most children. He is not good in typing so he type with one hand while looking at the keyboard. He starts typing and enters the spelling he remembers but he starts the word "asteroids" with the alphabet "e" instead of "a". He ends up entering the term "estroid" instead of asteroid in the search box. When he completes the spellings he looks up the screen and see the suggestions provided by Google and spots that the term he entered is also present so he becomes happy that he his spellings are correct. Now he presses enter and he is shown a long list of result. After examining the results for a while he realizes that these results are strange and not what he was expecting. He thinks he probably made mistake in spelling but he is also confused since he had selected the term from the suggestions provided by Google and after entering the query no further suggestions appeared in the "Did you mean" tag which made him believe that he had entered the correct term. This scenario is shown in the following figure.

Fig 3.1 Asjad entering the term "estroid" in search box and suggestions provided by Google

Above figure shows the case when Asjad is actually searching for "asteroids" but he is entering the spellings incorrectly and even suggestions provided are confirming his spellings so he mistakenly thinks that his spelling are correct but search results show that he is wrong. This situation can occur with anyone and especially with children when they think they are correct about the spellings and evidences also confirm their belief. So, when the results don’t turn out to be as expected, they think that there is no information available related to that particular topic and they become frustrated. They don’t realize that information is available and the fault is on their side. At this point if they don’t alter their keyword thinking that there is nothing wrong with the spellings or may be this thought doesn’t even come to their minds, then this situation can seriously hamper the beliefs of children about search engines and how they work.

Coming back to the example of Asjad, after he realizes that he is making mistakes in spellings he confirms the spellings from his textbook and notebook and catches the mistake he is making. So after a considerable effort he is finally able to enter the correct term.

Fig 3.2 Search results for the term "estroid"

Asjad thinks that after entering the correct spelling and getting the relevant results, the difficult part is over. But he is wrong. His actual and harder task is just starting. Now that he has entered the correct keyword and a long list of results is in front of him, a lot of questions are popping up in his mind e.g. why so many results, which of these are relevant and important for me, what points to focus and what to leave, and many other questions like these. Asjad realizes that getting a lot of results is very easy but comprehending and analyzing these results is very difficult task. He starts checking the results one by one and after a brief look he decides which page to keep and which to discard. He finds that not every result is relevant to his requirements. After scanning first two pages of results he thinks that he has gathered enough material and also the further he goes into results they are turning more irrelevant. He thinks why there are so many unrelated results and wishes that all the results very exactly according to his demands, which is off course a bit too ambitious to think. Now, Asjad moves on to the results he has short-listed for the essay assignment. He notices that there is no structured or standard format of the web pages. Few pages are written as simple text without any formatting of headings, while some pages are well organized and present the information in an ordered manner. So obviously he finds it easy to comprehend the pages written in a structures method, because it is simpler to single out the important concepts in the presence of heading since, headings obviously signify important things about the topic. On the other hand, in case of simple text he has to read the whole text, and it is totally up to him to think and decide which points are important and which are not. Somehow Asjad manages to prepare an essay after handling all the hurdles that came into his way.

Based on the above scenario we can clearly see the problems Asjad had to face during the task of searching and these problems are the same which have been highlighted earlier:

Spelling mistakes due to lack of vocabulary which in turn makes the task of keyword selection harder

Typing errors made due to less developed skills

Understanding and analyzing result set due to so many results and unstructured format of web pages

3.2 Kidzz Search – A Faceted Search Interface fir Children

We are now in a good position to introduce our interface based on the above scenario and the material covered earlier. We will try to justify how this interface caters to the various problems highlighted previously. I have named my interface as Kidzz Search.

Fig 3.3 Kidzz Search Interface

We can clearly see the different elements of the interface. We will now give the description of the functionality of these elements i.e. what do they do, their importance, etc.

3.3 Structure and Data Collection

A glance at the interface tells a few things about it. The page has two sections

Choose Options

Results

Choose Options section has few categories and check boxes under each category. Result section is empty when no selection has been made. We can see that Choose Options section has the main heading "Science" at the top. This is because all the pages which are stored and categorized are related to the subject science. Science is the subject which children start learning from early classed in their schools. It offers them the opportunity to explore and build upon their knowledge related to various domains of life since science involves everything from small parts of our body to the biggest secrets of this universe. So whenever a child learns a new and interesting concept from his class or hears about it from somewhere then he obviously wants to find more information about it. He goes to the Internet and starts searching about this particular topic using the search engine of his choice. So we have focused on the subject of science for the purpose of prototype creation and evaluation.

The next task is to apply the faceted search successfully. The essence of faceted search is to divide the domain in its attributes and perform search on the basis of those attributes. Here the domain of science is divided into four main categories Space, Planets, Human Body, Earth. These are by no means a complete and exhaustive list of categories. There can be a lot more other categories. These categories are only a sample from many categories. Each of these major categories is further subdivided into sub categories. Again there can be many other sub categories but only a small sample is taken for the prototyping level. These categories are made after analyzing different websites (kidsclick.org, nsdl.org) and few text books of children. The pages related to each category are found from different resources (mainly from kidsclick.org, nsdl.org) and simple web searching using different search engines Google, Yahoo, Bing.

The sub categories under each main category are shown as check boxes. Whenever a checkbox is checked the results related to that category are shown in the result section as shown in the image below

Fig 3.4 Solar System is checked and the results are shown

We can see the results in the Result section which are presented in a structure manner as heading at the top of the results goes like "Space > Solar System" which immediately helps in understanding that these results are related to Solar System in the Space category. We can also see that two further checkboxes have appeared. These are the facets which are used to filter the results. We will discuss these in details a bit later.

Fig 3.5 Kidzz Search when multiple options have been selected

What happens when user selects multiple options? This situation is depicted in the above figure. With each selected option its relevant results and corresponding headings appear in the Results section. User can check the results of any category from Results section. We can also see the relevant facets of each subcategory under it.

3.4 Facets and their Effect

We have seen that whenever a sub category is select few options appear under it. We have called them facets but the question remains what is the purpose of these facets? We touched upon them briefly as we said that these are used to filter the results of the sub category they belong. Now, we will discuss the working and effects of these facets in details.

As we know that Web consists of enormous amount of web pages which are heterogeneous in nature. To apply faceted search to webpages we need metadata, but web pages don’t contain enough metadata for the successful application of faceted search at a large level (Teevan et al. 2008). There are Meta Tags available in the webpages but they don’t contain enough information about a webpage. The only metadata we have about a webpage is its title or sometimes some keywords or description about the page provided by the author of the page. Other than that we know nothing about contents of a webpage in terms of metadata. It is not considered mandatory to provide metadata after designing the website. Developers are focused mostly on the design aspects of the process and they think their job is done as soon as the website is working. Also the task of providing accurate and quality metadata is not easy. It requires time and effort. Automatically generating quality metadata for webpages is also hard with the current state of the technology due to the diverse nature of the webpages and the inability to extract metadata which consistently applies to a large group of webpages (Teevan et al. 2008).

Keeping all these things in mind, the question arises how to apply faceted search to the webpages? If we analyze webpages it will become evident that there is no particular structure to the layout of the information. There are no standards to design webpages and everyone creates pages according to their preferences and requirements. Some pages only consist of written text without any formatting, headings, images. Some pages are written in a structured manner with proper headings and formatting. It is obviously easy to read and comprehend text which written in a structured manner as compared to the text which is written in a flow without any breaks or headings etc. It is even harder for children to read, analyze, and pick important points from plain text. Even with the pages written and presented in a formatted manner, it is difficult for children to digest the information quickly. They take time to read and reach the desired point in the text. Analytical and cognitive skills of children are not yet reached the point where they can cope with the information load easily hence the process of exploring information is slow in their case.

Here we have presented a brief analysis of the situation over the open web and problems which occur in understanding information due the unstructured nature of the webpages. The reason of this analysis was to reach a point where the hurdles in the path of successful application of faceted search become evident.

Coming back to the structure of webpages, if a webpage consists of heading then, these headings say something important about that particular topic. Headings are an indication of some important information coming up. Presence of headings in a page signifies that there is some important information under each heading otherwise there was no need to write that information in the separate space. Since we know the basic idea of faceted search i.e. to search the domain on the basis of its attributes, and we also know that open web consists of unstructured data and not enough metadata is available to apply faceted search on the open web, hence we can make use of the headings within a page and utilize it as metadata. So idea is to use headings found within webpages as facets to apply filters to the results within each section. It makes sense since headings are vital piece of information about text and hence can be counted as attributes of the domain. In Kidzz Search interface we have made use of this fact and applied faceted search on the basis of the headings found within the pages in our dataset. Whenever a subcategory is selected its results are shown in Results section but also the relevant facets associated to that particular subcategory also appear below the selected sub category.

Fig 3.6 Results after facets selection

In Fig 3.4 we can see the state of the system when the Solar System is selected and its corresponding results are shown in Results section. We can see Solar System has two facets; size and composition. The effect of selecting the option "size" is that it will filter the results in the Solar System section and only those results will be shown which have the heading "size" in their contents. Also a label will appear of the name "size" in front of all those results which have this heading. This label is clickable and it takes directly to place where the heading of "size" and its related material is written within the page.

Fig 3.7 Clicking "Size" takes directly to related the heading

Fig 3.6 shows that both the facets of Solar System have been selected and the results have been filtered accordingly in the Results section. Associated labels of each result have been shown in front of them. Fig 3.7 shows that when the label of "size" is clicked the page related to that label is opened and focus is shifted directly to the place where the heading of "size" is located. This is how Kidzz Search interface works.

3.5 Issues Handled by Kidzz Search

We will now discuss the advantages of using our designed faceted search interface, type of problems it solves, and how it helps children in finding information easily and quickly.

3.5.1 No Need for Query Input or Keyword Selection

One of the difficult tasks for children is finding proper keywords at the time of searching. This problem is solved by eliminating the need to enter query altogether. Since all the information can be searched on the basis of selecting appropriate categories so there is no need for the search box. So the time which is consumed in this activity is minimized hence using this searching interface is time efficient. All the children have to do is to find the appropriate category and the results related to those categories are show right away. Some would argue that how omitting keyword search counts as advantage. Selecting the relevant category is easier for children than thinking and reformulating queries again and again since children are more comfortable with recognizing objects than recalling the information (Borgman et al. 1995). Further help is provided to children by showing relevant facets which apply more filters and narrow down the results. Therefore, if sufficient information is covered in all the categories and options and it gives a good overview of the domain and directs towards the desired information then keyword search can be regarded as beneficial especially in case of children.

3.5.2 Typing and Spellings Mistakes are Avoided

Browsing based search engines have an added advantage, since children don’t have to type so they don’t make spellings and typing mistakes at all. Spellings and typing mistakes are other big issues for children other than keyword selection. They spend a lot of time in finding the required information because of the mistakes they make in spellings and typing, as once they enter the incorrect word and realize their mistakes after checking the results, they have to go back and do it all over again. So a lot of time is consumed in all these activities. Kidzz Search saves this time by providing the interface which is free from typing and spellings mistakes since there is no need for keyword search.

3.5.3 Hierarchical Representation Helps in Making Association and Sense of Results

It is very common to enter a query in search box and get thousands or millions of results. After that it is totally up to the searcher to make connections between the results and the query he entered. There is no explicit help provided by the search engines. Contemporary search engines now do try to provide some filters by dividing the results into different broad categories like Videos, Images, News, and etc. but when it comes to the actual contents of the results no filtration or categorization is provided. This situation becomes more serious when children perform the searching. Children are at such a level where they cannot process a lot of information as their cognitive skills are not matured enough. Search engines show the results in a set of 10 results per page but research has shown that even this information is more than they can absorb. As the amount of information grows, children have issues in establishing relevancy of results (Large and Beheshti 2000). In many cases they failed to do so and in turn believe that information does not exist. Kidzz Search suggests a solution to this problem as well. We have implemented the system such as that each category in Options section has a relevant category in the Results section and the each selection made to the right side brings the result in the left side and these results are only for the associated category. For the purpose of providing visual cues the title of each heading in the Results section has a title which shows the related category to which these results correspond. By arranging and presenting the information and results in this hierarchical way children are helped greatly since they can make strong association between the selections they made and the results they got in return. This representation also does not overload them with the information and they can easily understand what is happening.

3.5.4 Ease of Navigation Improves Satisfaction and Outcomes

Excess of everything is bad and it also applies to the amount of information related to a particular topic which is available to a person when he/she is looking for it. Excess of information is even worse in case of children. Children find a result after completing many difficult steps i.e. after overcoming the challenges of keyword selection, spelling and typing mistakes, and analyzing results for desired information. They finally reach a result after fighting hard against many deficiencies they have due to their undercook abilities. At this point what they want is that they are presented with their required information, but what they actually get in many cases is a lot of text. We discussed earlier that sometimes this text is formatted and many times unstructured and written in a sequence. This unformatted and unstructured text poses a great challenge to children. They actually have to read this whole text to find the exact piece of information they have been looking for. Now they don’t know whether this information is present in the current page or not but they give it a try. But children are just not good enough at such a small stage to grasp the ideas presented in the text. They read through the lines reading few words and skipping many and cannot focus on all the words. As a result, they usually end up assuming that this information is not available on the page even if it is present and they have missed it (Livingstone et al. 2005). Structured representation of text doesn’t also guarantee that children will locate the target position. If there are a lot of headings on the page, they might miss the intended heading, or they might not know the meaning of words which actually represent their desired information. Kidzz Search proposes a solution to this problem by defining facets, which are actually created by extracting the headings from the result pages associated to each category. Each facet selection creates a label in front of each associated result, and clicking this label takes directly to the place of heading. These facets are defined in simple language which is easily understandable for children. This way, children are saved from the efforts this whole process takes and which often results in the frustration which builds due to the failure to find information. As a result children become more satisfied as their efforts are now producing the desired outcomes and that too with the minimal efforts.

3.5.5 Facets Provide an Overview of the Domain

One important aspect of the faceted search is that it provides an overview of the domain just by looking at the facets and helps in developing an image of the domain in the mind. This in turn puts users in a better position to make decisions as he/she knows what to expect, and what sort of information is present. This is especially helpful in situations when users have no general idea about a particular domain so they face issues in defining queries. Facets provide the initial knowledge of domain to let users decide themselves what they are looking for. In case of Kidzz Search facets are the headings of the pages in the data set, and headings obviously say something about the contents of the page it belong. So children are helped here by providing a synopsis of the pages in the shape of facets and in this way they get a good idea of the information available on the pages.

3.5.6 Opportunities to Learn and Knowledge Building

Faceted Search facilitates learning and knowledge building. It indirectly builds the knowledge about the domain being explored and provides the opportunities to learn at each step. Facets are the attributes of the relevant domain which are chosen carefully to represent the domain in a best possible manner. So when people use a faceted search interface for information finding they have to closely observe the facets provided to accurately direct the system towards their required information. This process of correctly selecting the relevant facets ultimately helps users by building their knowledge about that domain, as they now know certain facts about that field which may have been unknown before using the faceted search system. Since facets in Kidzz Search come from the headings within pages from dataset, they contribute to the knowledge of children and add up to that knowledge.

3.5.7 Fewer Results Pose less Information Load on Children

Keyword based search engines work on the basis such algorithms which make them return many results. Now some would consider it a good thing but it is not so good particularly in case of children. Children examine only first page of the results in most cases and that too only first few results of the first page. If they don’t find the information they are looking for from these results they leave the task assuming that this information is not available. This behavior of children is justified since most relevant results are usually available on the first page and other pages rarely contain any valid result in most cases. Children will be comfortable with fewer but more accurate results which could satisfy their information needs. This is exactly the case implemented in Kidzz Search, as children get only few results as the outcome of each selection they make. These results are further pruned when facet selection is made. So each query results in fewer matches and children are not bombarded with the long list of results.

Article name: Childrens Use Of Internet And Search Engines Computer Science essay, research paper, dissertation

Make Assignments Great Again

24/7 customer support: science/80424-childrens-use-of-internet-and-search-engines.html