Information Design and Databases

Proximity is the idea of placing related elements close together and keeping unrelated elements apart, so groups are recognizable at a glance. Alignment is lining up positions and edges of elements, repetition is applying the same styling repeatedly to elements with the same role, and contrast is clearly differentiating levels of importance; none of these creates groupings.

A signifier is a cue that makes users notice how to operate something, such as a raised button that looks pressable or a door handle that invites pulling. Usability is the degree of ease of use, accessibility is the degree to which everyone can use something, and a persona is the archetypal user imagined during design.

Universal design is the idea of designing from the outset so that diverse users can use a product, rather than making separate versions for specific groups. The second option is the idea of preparing separate dedicated products, contrary to the definition; the third is design focused only on efficiency and the fourth on aesthetics, and neither expresses consideration for diverse users.

Alternative text conveys the content of images in words, so the information also reaches people using screen readers. Turning text into images prevents both reading aloud and font-size changes, so the first option is inappropriate; the second ignores individual differences in color perception; and the third fails to convey the content to people with hearing impairments; all are counterproductive.

Color perception varies from person to person, so combining cues other than color lets everyone distinguish the series. Red and green is a classic example of a hard-to-distinguish pair, small brightness differences make lines sink into the background and become unreadable, and subtle color differences cannot be told apart; all of those are inappropriate.

A pictogram is a pictorial symbol that conveys meaning independently of language; guidance symbols are specified in JIS Z 8210. Infographics are figures and illustrations expressing information of all kinds clearly, a signifier is a cue showing how to operate something, and a diagram is a drawing expressing relationships with lines and shapes.

Structuring means dividing information into meaningful units and showing the hierarchy with headings so things are easy to find. The second and third options are unorganized, leaving readers unable to grasp the whole. The fourth is a purely visual break that conveys no structure to screen readers and the like, so it does not amount to structuring.

Radio buttons allow only one of the lined-up options to be selected, and choosing another deselects the previous one. Check boxes toggle each item independently, allowing multiple or no selections; a text box is a field for free text entry; and a scroll bar moves the visible area; none of these suits single selection.

Check boxes toggle each item independently, allowing none or several to be selected. Radio buttons and a single-selection pull-down menu allow only one choice and do not fit the question. A text box takes free input, so wording varies and tallying becomes difficult.

A pull-down menu normally occupies only one line and, when opened, lets the user pick one item from a list, making it suitable for items with many options. Radio buttons take up space for every option, and check boxes are for multiple selection, so neither fits. A progress bar is a display showing how far processing has advanced, not an input component.

UX refers to the whole experience, including satisfaction and reassurance in addition to ease of use. The second option describes the UI, the user interface, which is just one element making up UX; the third is one metric of operating efficiency; and the fourth describes throughput; none is the definition of UX.

A VUI is a user interface using voice for input and output, allowing operation even when hands and eyes are occupied. A GUI is operated by pointing at shapes on a screen, a CUI by typing text commands; an API is a connection point through which software calls other software's functions, not a point of human interaction.

A check digit attaches a verification digit to the code, and errors are found by redoing the same calculation on the entered value and seeing whether it matches. A format check verifies the form, such as digit count and symbol layout; a sequence check verifies the ordering of data; and a range check verifies the magnitude of values; none is a mechanism that appends a digit.

A matching check compares the value against correct data prepared separately, such as a master, to confirm the value exists. A numeric check only looks for non-digit characters, a duplication check finds double entry of the same data, and a range check sees whether a value falls within upper and lower limits; none collates against a master.

A range check verifies that a value falls within the defined upper and lower limits, so it can reject 130. A numeric check only looks for non-digit characters and does not judge magnitude, a sequence check concerns the ordering of data, and a duplication check confirms whether the same data already exists; none can judge the size of a value.

Lossless compression restores data identical to the original upon decompression and is used for data in which not even one bit may change, such as documents and programs. Lossy compression raises the compression ratio by discarding part of the information and cannot be reversed. Streaming is a technique of playing while receiving, and archiving is bundling multiple files into one; neither is a category of compression method.

JPEG is a lossy format that reduces size by discarding information people rarely notice, and it suits storing photographs. PNG and GIF are both lossless and return the original image when decompressed. BMP basically applies no compression, storing each pixel's color information as is.

GIF is a still-image format that losslessly compresses up to 256 colors and supports background transparency and simple animation. The second option describes JPEG, which lossily compresses full color; the third describes uncompressed BMP; and the fourth describes PDF, which preserves layout; none is a characteristic of GIF.

PDF is an electronic document format that displays and prints while preserving layout, standardized as international standard ISO 32000. CSV is plain text with values merely separated by commas and has no formatting, HTML's appearance changes with screen size and settings, and MP4 is a video format holding footage and audio.

Streaming receives and plays in parallel, so the wait before viewing starts is short. The first option describes download playback, the third is not a way of watching video at all, and the fourth is wrong because sending without compression increases rather than decreases the traffic.

AR is a technology that overlays information on real-world imagery, reinforcing how reality is seen. VR blocks the real view and presents only an artificial space, MR merges the real and the virtual so they influence each other, and CG is the general term for creating images by computer.

MR is a technology that merges the real and the virtual, realizing an experience in which the two influence each other. The second option describes VR, which shuts out reality; the third is mere transmission of high-definition video; and the fourth is fabrication with a 3D printer; none amounts to merging real space with virtual objects.

4K means the horizontal pixel count is about 4,000, with 3,840×2,160 pixels being typical. 8K doubles both dimensions to 7,680×4,320 pixels. The first option is about frame count, the second about screen size, and the third about color count; none relates to the definition of 4K.

A DBMS provides concurrency control, exclusive control, along with confidentiality protection through access rights and recovery from failures. The first option is communications encryption technology, the third is debugging and verification work, and the fourth is the role of mail servers and mail software; none is a DBMS function.

The primary key identifies each row uniquely, so duplicate values are not allowed and null values are not allowed. The data type is not limited to numbers and may be a string, composite keys combining multiple columns are permitted, and there is no rule about column position, so the other options are not conditions.

A foreign key can hold only values that exist in the referenced primary key; this is called referential integrity. The second option is the work of a primary key or unique constraint, and the same product can of course be ordered many times. Deleting referenced rows is restricted, so the third is wrong, and the fourth has nothing to do with referential integrity.

Normalization is a design technique that eliminates the state of the same data being held redundantly in many places, preventing update anomalies where only one copy is fixed and the others diverge. Speeding up searches is the role of indexes, merging tables is the opposite idea from normalization, and encryption is a different matter; each has a different purpose.

An E-R diagram represents the structure of data through entities and the relationships between them, one-to-one, one-to-many, and many-to-many. The first option is what flowcharts and activity diagrams show, the second what organization charts and duty assignment tables show, and the third what network configuration diagrams show; none is the purpose of an E-R diagram.

Projection is the operation of extracting only the needed columns from a table, which is what happened here. Selection extracts the rows meeting a condition, a join connects two tables on a common column, and union is the set operation combining the rows of two tables of the same format; none of those narrows down columns.

Selection extracts only the rows satisfying a condition, leaving the column layout unchanged. Projection extracts only the needed columns, a join connects two tables on a common column, and difference is the set operation finding rows in one table but not the other; none of those is filtering rows by a condition.

Joining the two tables on their common column, department code, merges them into one table, and projecting the 3 columns employee number, name, and department name from it yields the desired table. No condition for filtering rows is given, so selection is unnecessary, and union and difference are set operations between tables of the same format, unusable for this purpose.

The rows with unit price 300 or more are the 2 rows Mug at 800 and Towel at 500, and the projection narrows the table to the 2 columns product name and unit price, giving 2 rows and 2 columns. The first option is the size without narrowing the columns by projection, the second without narrowing the rows by selection, and the third has the wrong number of matching rows.

In the join, rows with matching department codes are linked, and each of the 3 employee rows matches exactly one department, so the result has 3 rows. D3, General Affairs, to which no employee belongs, does not appear in the result. 9 rows would be the count if every combination were formed, and there is no basis for 4 or 6 rows.

Exclusive control has the process that starts updating first place a lock on the data, making other processes wait and preventing inconsistency. Backup saves copies in preparation for failures, normalization is a design technique eliminating duplication, and an index speeds up searches; none of them controls simultaneous updates.

A deadlock is a state in which transactions wait for each other to release the locks on resources the other holds, and it will not resolve if left alone; it is cleared by forcibly rolling back one of them. The first option is merely lengthened waiting, the second is a media failure, and the fourth is data duplication; none is a standstill caused by mutual waiting.

Rollback uses the before-image log to return the database to its state before the transaction started. Commit finalizes updates, a checkpoint writes out and records contents as a starting point for recovery, and rollforward reproduces updates up to just before a failure using the after-image log; both the direction and the purpose differ.

Recovery from a media failure restores the backup and then performs rollforward, reproducing the subsequent updates with the after-image log. The first option goes in the backward direction and cannot recover the latest state. The third is not a realistic means of recovery, and the fourth is table structure design work unrelated to restoring data.

Data mining digs out tendencies and relationships from large volumes of data using statistical and machine learning techniques. A data warehouse is a database accumulating data chronologically for analysis, a data mart is a subset carved out of it for a specific use, and data cleansing is the work of fixing inconsistent notation and errors to raise data quality.

A data lake stores raw data as is, whether structured or unstructured, and the data is processed for analysis when it becomes needed. The second option is a data mart, the third the tables obtained through normalization, and the fourth a BI tool dashboard; none is the definition of a data lake.

NoSQL is the general term for databases that are not bound to the table format and handle large, diverse data fast in structures such as key-value and document types. The first option describes SQL itself, the second a design technique, and the third a read-only access right setting; none is the definition of NoSQL.