Quinterest: Hassle-free Database Search

[rohitx15]

http://www.quinterest.org



About Quinterest - http://www.quinterest.org

Quinterest is a quiz bowl-oriented website dedicated to helping players everywhere sharpen their skills and increase their knowledge. The name derives from the combination of “quiz bowl” and “interest.”

This project was coded by me, Rohit Lalchandani. The database was constructed through a collaboration of my close friend, Joshua Duncan, and I.

---------------------------------------------

Currently, Quinterest is simply a very elegant, streamlined database search engine. However, with time, there is much more to come! Certain aspects were taken into consideration when the search engine was created that would inprove upon user experience.

For example:
-The search engine is designed to be clean and easy-on-the-eyes, allowing users to focus more overall and have a better experience, all thanks to Twitter Bootstrap.

-When viewing results, there is no pagination (unlimited results per page). This allows users to identify clues from individual questions more easily. In addition, when searching, the page does not refresh which saves your selected search parameters and reduces bandwidth usage.

-There is the ability to further specify search queries with the option to search by difficulty. This allow users to focus on questions related to their level of quiz bowl.

---------------------------------------------

Technical Stuff (Most of you can just skip this if you wish and just head to the website: http://www.quinterest.org)

-The site's front-end is constructed using a custom version of the Bootstrap framework. The Twitter Bootstrap framework keeps designing the website simple but also allows for a beautiful design.

-The site consists of HTML, CSS, Javascript, and PHP.

-The database is stored in MySQL which is very efficient as processing queries quickly and efficiently. The database is scalable and more and more questions are being added everyday. ('Bonus' search is coming soon!)

-The search parameters can be set to search by difficult: All, HS, MS, or College.

-Included on the main page is a link to a copy&paste/printer friendly version of the site which removes all the formatting and CSS, leaving a barebones quiz bowl database search.

---------------------------------------------

Contact

If you have any questions, comments, concerns, suggestions, or would like to report a question with an error, it is encouraged that you send an email to info@quinterest.org

[Corry]

Maybe Quizbowl DB's ridiculously wonky search feature has just substantially lowered my standards, but I think this is really great. I mean, really, REALLY great. I'll be using this for sure. Thank you so much!

[High Dependency Unit]

It is A LOT better than the qbdb search. This one actually searches using all of the words in the search.

[Corry]

I've been messing around with Quinterest for the past hour or so, and there are a few things I want to note:

• A lot of the questions in the college section are preemptively cut off. Like, nearly a quarter of them. This problem doesn't happen the high school section. Example: http://i.imgur.com/7sJNvsk.png
• Can we get an option to filter by categories? The search engine already lists categories for each question, but there's no way to sort through one category or another.
• There seem to be a few duplicate questions here and there. Two examples that I found: http://i.imgur.com/kP4Afmu.png and http://i.imgur.com/L4PiOGP.png
• An option to search through the text of the question (rather than just the answer line) would be awesome.

Anyways, keep up the good work.

[rohitx15]

I've been messing around with Quinterest for the past hour or so, and there are a few things I want to note:

• A lot of the questions in the college section are preemptively cut off. Like, nearly a quarter of them. This problem doesn't happen the high school section. Example: http://i.imgur.com/7sJNvsk.png
• Can we get an option to filter by categories? The search engine already lists categories for each question, but there's no way to sort through one category or another.
• There seem to be a few duplicate questions here and there. Two examples that I found: http://i.imgur.com/kP4Afmu.png and http://i.imgur.com/L4PiOGP.png
• An option to search through the text of the question (rather than just the answer line) would be awesome.

Anyways, keep up the good work.

Thank you. Your concerns and feedback are well-appreciated. We are aware that the database had errors and duplicates. We are hard-at-work to try and clean up the database and continue to expand it (we have also began working on a bonus search section). Anyways posts like yours are exactly what we are looking for in order to pinpoint errors and certain questions that need to be fixed. In addition, we would truly appreciate it if users would email specific problems to info@quinterest.org. This really helps us as we continue to polish the database and free it from errors as much as possible.

Also, as for the extra search parameters such as category, tournament, searching through question, etc., I am very very new to programming and creating projects such as Quinterest in general. So, as I continue to learn, I will continue to add features and extras. But to answer your question, yes, that is on the list of things we would like to add eventually. :)

[NikhilSethi]

Just wondering, but what sets does this include? Does it just contain the questions found on quizbowldb and protobowl?

Thanks for making this! Searching was getting annoying with QBDB's issues going on.

[rohitx15]

Just wondering, but what sets does this include? Does it just contain the questions found on quizbowldb and protobowl?

Thanks for making this! Searching was getting annoying with QBDB's issues going on.

Thank you for your concern. A list of all the packets that are contained in the database can be found here: http://textuploader.com/?p=6&id=hM87U

[Adventure Temple Trail]

I've noticed that several questions cut off at the first character with an accent mark (e.g. the question ends abruptly at "Melqu" or "[accept Cien a" on a search for One Hundred Years of Solitude).

[rohitx15]

I've noticed that several questions cut off at the first character with an accent mark (e.g. the question ends abruptly at "Melqu" or "[accept Cien a" on a search for One Hundred Years of Solitude).

I will be pushing out a small update later tonight (probably a few hours) that corrects all the questions that have been cutshort. I found an error in some of my scripting which I've fixed now. I just need to construct and reupload the new database.

As for category searching, I've created a working model, although it is very crude. I'll probably add it to the site in a few days worth.

Thanks.

[rohitx15]

I've been messing around with Quinterest for the past hour or so, and there are a few things I want to note:

• A lot of the questions in the college section are preemptively cut off. Like, nearly a quarter of them. This problem doesn't happen the high school section. Example: http://i.imgur.com/7sJNvsk.png
• Can we get an option to filter by categories? The search engine already lists categories for each question, but there's no way to sort through one category or another.
• There seem to be a few duplicate questions here and there. Two examples that I found: http://i.imgur.com/kP4Afmu.png and http://i.imgur.com/L4PiOGP.png
• An option to search through the text of the question (rather than just the answer line) would be awesome.

Anyways, keep up the good work.

I've noticed that several questions cut off at the first character with an accent mark (e.g. the question ends abruptly at "Melqu" or "[accept Cien a" on a search for One Hundred Years of Solitude).

As promised, the majority of the question errors are now corrected. I updated the database, and there shouldn't be any more questions that are cutoff. As for duplicates, I am working hard to find those, but at least every question's text is fully there now. Also, category search, question/answer search, and improved search overall are coming soon.

[rohitx15]

I put out a pretty sizeable update this morning (spent the whole night figuring this out). Anyways, there is now the ability to narrow your search by category as well as difficulty. In addition, almost all of the errors in the database have been fixed (their might be a question here and there that has an error). Every question has an assigned ID and that is now shown in the top right corner of the question box. This makes it easier for users to email me (which I deeply encourage) questions that have errors and need to be fixed. Also, now, when a user searches, there is no longer a page refresh. The results appear below the search without the page have to reload.

Overall, I'd say this was much needed--especially ridding the database of cutoff questions and answers. Also, keep looking out, there is much more to come (Bonus Search). :)

[theMoMA]

Looks really cool, and helpful for writers looking to see how recent tournaments have asked about certain answers.

A few suggestions:

(1) Would it be possible to add basic search commands, such as quotation marks for exact phrases or a dash for eliminating certain words from the results?

(2) As others have mentioned, the option to search in the question text (not just the answer line) would be excellent.

[rohitx15]

Looks really cool, and helpful for writers looking to see how recent tournaments have asked about certain answers.

A few suggestions:

(1) Would it be possible to add basic search commands, such as quotation marks for exact phrases or a dash for eliminating certain words from the results?

(2) As others have mentioned, the option to search in the question text (not just the answer line) would be excellent.

A 'search type' option has now been added. Thank you for your suggestions. I'll try my best to implement as much as I can.

[Eddie]

"Philosophy" is spelled as "Phylosophy" in the "Category" drop-down.

[rohitx15]

"Philosophy" is spelled as "Phylosophy" in the "Category" drop-down.

Haha. You are correct. It has been fixed.
:)

[rohitx15]

Sorry guys. I've been on vacation for a few days now and have been unable to access a computer. I have received and replied to all the emails sent to me and corrected all the database errors. Keep them coming! They really help. Thanks again for all the emails.

More features are coming soon. Keep a lookout!

[rohitx15]

Hello, everyone! I have updated Quinterest with a "Report an Error" button!

It is now possible to report a database error without the need to use an email client. I hope this will encourage users to report incorrect questions more since it removes a lot of the hassle of creating and sending an email.

Anyways, look for more soon.

Rohit Lalchandani

[nadph]

Is the code for this posted somewhere, e.g. GitHub?

[sbfromcopley]

I have been experiencing a lot of success using this site over quizbowl db, having all the questions on one page is great for looking things up. One thing I have been experiencing a lot of is repetition of questions A LOT. For example, if you look up Zoroastrianism, you will get 35 results! Wow, so many questions, right? But when you read through them you realize it is just a few questions being repeated 3-4 times for some reason. The website is still very helpful, the repetition is just a blow at the efficiency of the site. It would be very helpful if you could get that fixed.

P.S. I also reported problems for all of the repeats on Zoroastrianism so you can take care of that specific answerline.

[rohitx15]

I have been experiencing a lot of success using this site over quizbowl db, having all the questions on one page is great for looking things up. One thing I have been experiencing a lot of is repetition of questions A LOT. For example, if you look up Zoroastrianism, you will get 35 results! Wow, so many questions, right? But when you read through them you realize it is just a few questions being repeated 3-4 times for some reason. The website is still very helpful, the repetition is just a blow at the efficiency of the site. It would be very helpful if you could get that fixed.

P.S. I also reported problems for all of the repeats on Zoroastrianism so you can take care of that specific answerline.

My apologies for the inconvenience. I am trying my hardest to clean the database, and the error reports are extremely helpful. As for duplicates, a lot of the duplicates come from HSAPQ reusing their toss-ups in difference tournament, but some of them also come from being direct duplicates from the same tournament. I am removing more and more of these duplicates each and every day. I can't think of a way to find all duplicates in a quick fashion so this will have to do for now, but if a more experiences programmer would like to help me out a bit, I will gladly accept their advice and direction. Nonetheless, I am improving Quinterest constantly and if others would like to provide advice, please feel free to send a message to my hsquizbowl inbox or email info@quinterest.org.

Thanks, once again.

EDIT: After tinkering around a bit, I found a way to find all duplicates in the database. I still have to manually go through and delete the duplicates, but I no longer have to go and search for them. This probably speeds things up by over half the time it would have taken!

[heterodyne]

Just wondering, although this is obviously a low priority, would it be possible to make the database searchable by the tournament the question came from?

[rohitx15]

Just wondering, although this is obviously a low priority, would it be possible to make the database searchable by the tournament the question came from?

As per requested, I have added the ability to search by tournament as well now. If you have any other suggestions or concerns, please feel free to express your opinion. I take all feedback seriously and consider all discussion openly.

More later.

[rohitx15]

I know this may be a little late, but if anybody would like to keep updated with the progress and work on Quinterest, I have a blog at http://www.rohitlalchandani.blogspot.com.

I will periodically post information about tweaks and changes to Quinterest as well as features planned for the future, if anybody is interested.

[rohitx15]

I just added a small update (this puts Quinterest on version 0.9.2) that even though it is very minuscule, I think many will appreciate it. Now, when viewing search results, there is a tag beside each result that indicates what number result it is. This is very helpful when there are a lot of results so the user doesn't get lost when viewing them.

Keep on the lookout for more features coming soon!

[rohitx15]

After running some stuff through MySQL, ALL the duplicates should be removed now. If, by some off chance, anybody sees a duplicate in the database, please send an error report, but I don't think anybody should. This should probably please quite a few people (myself included) as seeing the duplicates was starting to get extremely annoying. New features are coming soon.

[rohitx15]

http://rohitlalchandani.blogspot.com/2013/07/something-big-is-coming.html

[Corry]

http://rohitlalchandani.blogspot.com/2013/07/something-big-is-coming.html

Sounds fun.

[rohitx15]

Small update for Quinterest (now version 0.9.3):

When viewing results in Quinterest, users may have noticed that many questions contain double quotes ("") where there should only be a single quote ("). This issue has been fixed. Please continue sending error reports! Every single one of them that has been sent has been fixed, and they make it super easy for me to find errors.

[rohitx15]

http://rohitlalchandani.blogspot.com/20 ... pdate.html

[rohitx15]

http://rohitlalchandani.blogspot.com/20 ... st-10.html

[Eddie]

Is there a way to search for two separate strings in the same question, without searching for both strings together? For example, searching for "apples" and "oranges" but not necessarily "apples oranges."

[Ubermensch]

Although Quinterest does make finding questions easier, the number of questions that come up on a particular answer line is significantly less than on Quizbowldb. For example:

Things Fall Apart Tossups on quizbowldb: 17
Things Fall Apart Tossups on quinterest: 8

Grapes of Wrath Tossups on quizbowldb: 8
Grapes of Wrath Tossups on quinterest: 3

Seeing as my purpose for using any question database is to gather stock clues, I will continue to use quizbowldb as it provides a larger sample size. Although, the search function on quizbowldb is annoying, there are ways to circumvent it. Unfortunately, there is no way to circumvent the fact that there are not as many questions. Therefore, I suggest adding more packets to the database from the packet archive.

[AKKOLADE]

Although Quinterest does make finding questions easier, the number of questions that come up on a particular answer line is significantly less than on Quizbowldb. For example:

Things Fall Apart Tossups on quizbowldb: 17
Things Fall Apart Tossups on quinterest: 8

Grapes of Wrath Tossups on quizbowldb: 8
Grapes of Wrath Tossups on quinterest: 3

Seeing as my purpose for using any question database is to gather stock clues, I will continue to use quizbowldb as it provides a larger sample size. Although, the search function on quizbowldb is annoying, there are ways to circumvent it. Unfortunately, there is no way to circumvent the fact that there are not as many questions. Therefore, I suggest adding more packets to the database from the packet archive.

I, too, am outraged that a search engine that's existed for a month doesn't have as many questions, despite the fact that it's actively working at expanding its databases.

[rohitx15]

Although Quinterest does make finding questions easier, the number of questions that come up on a particular answer line is significantly less than on Quizbowldb. For example:

Things Fall Apart Tossups on quizbowldb: 17
Things Fall Apart Tossups on quinterest: 8

Grapes of Wrath Tossups on quizbowldb: 8
Grapes of Wrath Tossups on quinterest: 3

Seeing as my purpose for using any question database is to gather stock clues, I will continue to use quizbowldb as it provides a larger sample size. Although, the search function on quizbowldb is annoying, there are ways to circumvent it. Unfortunately, there is no way to circumvent the fact that there are not as many questions. Therefore, I suggest adding more packets to the database from the packet archive.

I apologize if you've had any inconvenience with Quinterest, but last I checked, searching Grapes of Wrath on quizbowldb.com yields 7 results that are actually about Grapes of Wrath and 28 results that have other answer lines. Searching Grapes of Wrath on Quinterest yields 7 results all related to Grapes of Wrath. Results are similar for Things Fall Apart as well. If you take notice, quizbowldb has a database of 33094 questions (some of which are duplicates since HSAPQ recycles their questions) and Quinterest has 29519 (none of which are duplicates). And then there is also the new bonus search now, which consists of 1034 bonus questions. I apologize, but your examples are not accurate.

I hope that Quinterest can be of some use to you in your future studying for quiz bowl.

If you have any suggestions or concerns, feel free to post here or send an email to info@quinterest.org.

[Matt Weiner]

some of which are duplicates since HSAPQ recycles their questions

To make this perfectly clear: Like all other vendors, HSAPQ offers questions in custom formats upon the request of a tournament. Unlike all other vendors, we make our questions available for free at the end of each year, including sets using the same base of questions in multiple formats. We do not "recycle questions" across years or across tournaments that the same team will be playing. The duplicates are from what is essentially the same meta-set--e.g. "the questions from which VHSL States and OAC were drawn in 2013"--formatted in different ways for use in different areas by different teams in the same year.

With some fairly rare exceptions, when this is done, the bulk of the question text remains unchanged, so you can filter these out by searching for tossups with the same first sentence or so on.

[rohitx15]

some of which are duplicates since HSAPQ recycles their questions

To make this perfectly clear: Like all other vendors, HSAPQ offers questions in custom formats upon the request of a tournament. Unlike all other vendors, we make our questions available for free at the end of each year, including sets using the same base of questions in multiple formats. We do not "recycle questions" across years or across tournaments that the same team will be playing. The duplicates are from what is essentially the same meta-set--e.g. "the questions from which VHSL States and OAC were drawn in 2013"--formatted in different ways for use in different areas by different teams in the same year.

With some fairly rare exceptions, when this is done, the bulk of the question text remains unchanged, so you can filter these out by searching for tossups with the same first sentence or so on.

My apologies if my words are misinterpreted. You are correct. That is what I was trying to say. I didn't mean recycle in that they use the same questions year after year. Thank's Matt for the clarification.

[Ubermensch]

I did this comparison the day before posting my comment. I see now that the numbers have changed. Regardless, that subtle difference in questions ultimately is the reason for my continued use of quizbowldb. In the case that searching for something in quizbowldb becomes tedious to such an extent that finding an extra question or two does not seem worth the search, I will most definitely be using quinterest.

I should also add that the purpose of my comment was to voice what I think is a common concern among those who have similar intentions for using these databases. The fact that you have created a database that has such a large quantity of questions in such a well-organized system is much-appreciated.

On the subject of the HSAPQ questions that were used for multiple tournaments, there are often-times differences between these questions where a clue is replaced for another clue. I understand why you have chosen not to include them in your database.

[agarg]

I just found Quinterest and I must commend you on such a great database Rohit! This will likely soon become my primary method for studying!

[rohitx15]

I just found Quinterest and I must commend you on such a great database Rohit! This will likely soon become my primary method for studying!

Why, thank you! I really appreciate the compliments. And don't worry, I know there are still some minor problems with the database with questions and whatnot, but I am working to get those fixed. I've been really busy lately and haven't had nearly as much time to work on Quinterest as I've had before. No worries, though. Everything will be fixed soon enough. I am also working on adding more questions--tosups and bonuses.

I hope you enjoy Quinterest. It'll only get better from here.

[mithokie]

Would you consider an option to print out tossup and bonus results all together as the result of one search? I was thinking along the lines of 3 options

1. tossup results only
2. bonus results only
3. both tossup and bonus results.

[JKHtay]

Would it be possible to see which topics have the most tossups about them from a particular category? Something like this?

[Galadedrid Damodred]

First off let me say that this is a fantastic application and I very much look forward to further improvements. Here are some comments I have on the current build:

1. Echoing what others have said upthread, having some metacharacters for search options would be great. At the moment I am forced to use "this" as a crude method of searching all questions in a category. Also, when I searched "thor" under Mythology, a bunch of non-Thor-related questions came up because they had the word "author" in them. So you can see why being able to do a more advanced search would be a major upgrade.
2. When reporting an error, it would be really nice if the ID of the question being reported were auto-filled in. Right now you have to remember it or write it down before clicking the report button.
3. After selecting a difficulty, only the tournaments at that difficulty level should show up in the rightmost dropdown menu. I would rather not wade through a bunch of middle & high school listings every time I am looking for a college or open tournament.
4. Is it possible to have further subdivisions within the distribution (e.g. being able to choose Physics instead of just Science), or is that asking for too much? Also, the list freak in me would prefer that the places of Religion and Trash be swapped in order to follow the standard ACF ordering.

I and others really appreciate all the work that has gone into Quinterest and it looks sure to become quite popular in the near future, if it hasn't already. Awesome job, Rohit!

[vinteuil]

4. Is it possible to have further subdivisions within the distribution (e.g. being able to choose Physics instead of just Science), or is that asking for too much?

As much as we'd all like this, it's almost impossible to implement—first, you'd have to recategorize the whole database, second, there would be a huge number of "mixed" categories that would blow up the size of the whole thing, third, sometimes it's really hard to tell the difference between, say, chem and bio or chem and physics, if there is one—so that would slow down the expansion of the database immensely.

[Galadedrid Damodred]

4. Is it possible to have further subdivisions within the distribution (e.g. being able to choose Physics instead of just Science), or is that asking for too much?

As much as we'd all like this, it's almost impossible to implement—first, you'd have to recategorize the whole database, second, there would be a huge number of "mixed" categories that would blow up the size of the whole thing, third, sometimes it's really hard to tell the difference between, say, chem and bio or chem and physics, if there is one—so that would slow down the expansion of the database immensely.

I figured as much, and I'll of course be happy with whatever level of sophistication is possible.

[quizbowllee]

I'd like to take the opportunity to say that Quinterest.org is one of the most useful tools I've ever seen in my time coaching. I'd like to see it expanded with more questions. I'm not very knowledgeable of how these things work, but I'd be more than glad to help out in any way possible in order to get more tournaments added to the database. I'd particularly like to see HS questions from the past year or two. My kids have started using this site a lot and I'm seeing great improvement.

Thanks for the hard work that I know is put into such a huge project.

[zman147]

Just wondering why PACE NSC and LIST only have first round tossups entered onto Quinterest instead of the whole packet. This might be a problem with other packets as well, I didn't check, but is it on purpose?

[High Dependency Unit]

Would it be possible to have a study feature for bonuses similar to the one for tossups?

[Corry]

Would it be possible to have a study feature for bonuses similar to the one for tossups?

Actually, that already exists.



However, not very many bonuses have been parsed yet. It would be nice if the feature were expanded.

[High Dependency Unit]

I mean a study feature that retrieves random bonuses, not a specific database search.

[The ChatSack Triple-Play]

4. Is it possible to have further subdivisions within the distribution (e.g. being able to choose Physics instead of just Science), or is that asking for too much?

As much as we'd all like this, it's almost impossible to implement—first, you'd have to recategorize the whole database, second, there would be a huge number of "mixed" categories that would blow up the size of the whole thing, third, sometimes it's really hard to tell the difference between, say, chem and bio or chem and physics, if there is one—so that would slow down the expansion of the database immensely.

Speaking as someone who came very close to putting out a database of his own before being one-upped by scary internet Jerry and his Ph.D., it is in fact possible. You'll want to look into naive Bayes classifiers (which yielded me a fulfilling power at PACE) and/or this paper. You can retrieve and parse all of ACFDB, which have been classified by hand into subcategories, and are generally good material, to use as a corpus. Yes, common links and ambiguously classified questions are things, specifically sizable wrenches in the works, but you can't have it all with something as squishy as this unless you want to do it yourself.