=Paper=
{{Paper
|id=Vol-2304/00030053
|storemode=property
|title=Towards a Logic lab on the Cloud: Desktop and Mobile Sessions
|pdfUrl=https://ceur-ws.org/Vol-2304/00030053.pdf
|volume=Vol-2304
|authors=Debora Rodriguez-Aguilar,Juan C. Acosta-Guadarrama,Víctor Morales-Rocha
}}
==Towards a Logic lab on the Cloud: Desktop and Mobile Sessions==
https://ceur-ws.org/Vol-2304/00030053.pdf
Towards a logic lab on the cloud: desktop and mobile
sessions?
Debora Rodriguez-Aguilar, Juan C. Acosta-Guadarrama[0000−0002−4837−1578] , and Víctor
M. Morales-Rocha[0000−0001−8213−400X]
National Laboratory for Information Technologies, LANTI, Mexico;
http://www.lanti.org.mx
Autonomous University of Juarez, Institute for Engineering and Technology, JUÁREZ,
CHIHUAHUA, 32310, MEXICO
al132081@alumnos.uacj.mx, {juan.acosta,victor.morales}@uacj.mx
Abstract. This case study proposes a novel system of the Internet as a service
for Answer Set Programming (ASP) solvers, keeping open sessions for registered
users, in such a way they can work ASP knowledge bases anywhere, without
having to download, compile, install and configure the solver. The technique can
exploit the cloud and mobile potentials, such as processor and memory manage-
ment, GPS (Global Positioning System), accelerometer, thermometer, heart rate
sensor, touchscreen and gestures, voice, images, fingerprints, signatures, phone
calls, wristwatch, proximity sensor, and so on. It can also get benefited from cloud
computing and data mining potentials, such as massive computing and memory
distribution, big-data collection and analysis, centralized work, work sessions,
scalability, concurrent teamwork, and more. We started this proposal with the
aim to offer an alternative to the everyday use of DLV and CLINGO ASP solvers,
readily available on devices that everybody uses, like browsers and mobile de-
vices. This time we focus on setting up a global service, the desktop browser, and
mobile interfaces. We also provide a methodology of incremental development,
consisting in showing work progress into phases of analysis, modeling, design,
and implementation of the solution. In the following, we introduce the problem
of computing ASP knowledge bases online and background. Later we add a brief
survey on similar works. Then we present our proposal. Finally, we conclude with
a discussion and future work.
Keywords: Answer Set Programming solvers · cloud computing · web program-
ming · mobile devices · data mining · usability.
1 Introduction
Answer Set Programming (ASP) is a consolidated declarative programming paradigm
first proposed back in 1988 [11]. It provided a foundation to study and experimentation
in several investigations, such as the Frequency Assignment Problem, FAP [9], Mod-
eling of Biased Decisions [20], to mention a few. They all make use of ASP software
?
This work has been supported by mainly supported by The National Council for Science
and Technology, CONACYT, 280712, Consolidación de Laboratorios Nacionales. We are also
grateful to the Autonomous University of Juarez, Mexico.
54 D. Rodriguez-Aguilar et al.
as their core engine. In the same way, there exist development projects to implement
generic inference engines under the ASP language, called solvers. A solver is a mech-
anism that calculates conclusions in ASP semantics, that is to say, that, from specific
syntactic rules, it admits well-formed logic sentences, interprets and evaluates them
from the knowledge base (KB), and finally gets out the models that satisfy the KB
to solve a problem. At present, there are several solvers or inference engines, such
as Smodels, PSmodels, and others. However, this case study focuses on DLV [4] and
CLINGO [10,21] only, both world-renowned by researchers from around the world. The
popularity of both systems is due to their multiple applications in areas such as mathe-
matical logic, knowledge representation, and reasoning, as mentioned before. However,
the original use of an inference engine is limited to being used by a command-line
interface—CLI.
The problem with traditional command-line applications is that one needs to down-
load them, sometimes to compile them, to install them and to run them on the terminal.
For that, most of the times one needs unique skills in programming and operating sys-
tems administration, like Linux/UNIX. It is not just old-fashioned and impractical, but
also out of the scope of today’s Artificial Intelligence (AI) demands. Today’s AI ap-
plications demand the use of online services, concurrency, mobile devices, lots of data
collection, merging of different heterogeneous sources of information, big-data process-
ing and interpretation, which are not an easy task. So, one of the first steps to tackle is an
adequate modern graphical user interface (GUI) for the web apps and mobile devices.
Not only is it practical and novel but necessary for today’s demands. Our proposal con-
sists of a web front-end prototype for those two solvers and more, a portal that admits
both registered or unregistered users, concurrency, cloud computing, saved sessions for
registered users, and a native Android mobile app to access the cloud service. The first
aim of the proposal is to provide a modern ASP solvers environment both for scientists
and students, for their experiments and teaching on the formal system. However, its use
shall be not limited to that. If we implemented an adequate front-end, it would open
up new potentials and branches of research, like scalable cloud computing; and pro-
cessing of big data collected by mobile devices. The integration of mobile devices with
ASP is not luxurious. It permits other nice features desktop ASP applications lack, like
portability, and the collection of detailed information to make inferences from, such as
frequently visited places, road routes and (path) planning, health monitoring, shopping
and sports habits, voice processing and image processing, phone call classifications,
text messaging, and scheduling, to mention a few. The collection of such data would be
possible due to the different input devices a cellular phone has, like wristwatch apps,
proximity, the global-positioning system (GPS), fingerprint, touchscreen, gestures, ac-
celerometer, thermometer, heart-rate sensor, microphone and camera, and so on. It is
about exciting new theory and applications of ASP, human-computer interaction, Nat-
ural Language Processing, Multi-Agent Systems, and Ambient Intelligence; towards
finding patterns in human behaviors and services.
Our proposal presents LogicSite, a multi-platform application prototype both for
desktop computers and Android mobile devices. It is intended for users of ASP solvers,
who wish to focus on their semantics work rather than dealing with downloading, in-
stalling, configuring and typically running text-based interface ASP solvers. What is
Towards a logic lab on the cloud: desktop and mobile sessions 55
more, online users do not even have to worry about a fast parallel processor or memory
issues. In this paper, our main areas of study are the implementation of ASP solvers on
the cloud, as well as a native mobile app, which can exploit the benefits of such technol-
ogy. Having such services fully-implemented and available shall keep their users from
intricate details of managing local resources to download, compile, install, configure,
and run them.
2 Background
As far as we can tell, at present, the existing projects associated with the use of ASP
inference engines are intended for web development. Evidence of this assertion is avail-
able with the following applications:
– LoIDE [6,12,13] is an integrated development environment (IDE)1 of type web-
based—software for the web. Its objective is to offer an advanced and modular web
editor for ASP logic languages (DLV and CLINGO). Among its functions are the
execution, the processing of logic statements from an external file2 , and an option
to download the software to run it locally. Also, it includes functions for the inter-
face such as keyboard shortcuts, selection of filtering options and customization—
design and appearance.
– LogicLab [1] is a web app that aims to provide online-running inference engines
in a graphic browser. So, users need to focus on their work only, rather than other
details like how to download the solvers, how to install them, how to configure them
and how to run them on a text terminal. The ASP solvers available in LogicLab are
DLV and CLINGO, among others, both with logic program processing parameters.
– Running Clingo [22] is an ASP tool that, as its name says, implements the CLINGO
inference engine. Among its functions, we can highlight the reasoning mode, in
which one can select an option from the four existing ones—default, brave, cau-
tious and enumerating all. Besides, they include a list of ready-to-run examples.
– EmbASP [6,7] is a framework for the integration of logic programming in external
systems for the general-audience applications. It offers two implementations (in
Java and Python), in addition to providing DLV, CLINGO and DLV 2.0. Mobile
apps like DLVfit [6,7,8], GuessAndCheckers [6,7,3], DLVEdu [6,7], and Connect4
[6,7], they have used this framework for logical deductions that they generate and
execute internally.
Despite the existence of these applications, the solution they offer individually is not
yet satisfactory; this is because one application has functions that another does not and
vice versa. This table shows the main features that, in our view, we consider viable or
acceptable and desirable for an online laboratory to provide modern service of ASP
solvers.
These projects display logic calculation tools in a modern graphical environment.
The latter is something new because the everyday use of ASP solvers is through an
old-fashioned command line interface, and we had talked about its cons above.
1
An IDE is a software system that allows code editing, compilation and debugging.
2
The type of files that it allows is JSON—JavaScript Object Notation, a format for the exchange
of data over the network.
56 D. Rodriguez-Aguilar et al.
3 Preliminaries
In this section, we introduce some basic concepts, needed to underpin our proposal. The
main ideas here are portability, scalability, and usability, ans security, which state a sort
of principles to meet by adequate applications for modern technologies, in our view.
Later, we introduce some simple ASP background, where, in this paper, we assume the
reader is familiar with its syntax and semantics, easily accessible in the literature.
3.1 Usability
Jakob Nielsen [17] explains, to begin with, that “usability is a fundamental character-
istic for customers to visit, take advantage of and return to a website again, increasing
the success of their Internet experience.” As a result, we define usability as the way in
which users interact with a website, application or mobile app. The usability of systems
System Usability
Attribute Definition Measurement Method
Ease of Learning It is the minimum time required for Required time for a beginner user to
the user to learn of the functions of reach the level of expertise.
the application and to increase their
productivity.
Efficiency Ability to perform or fulfill a task The time required to perform a task.
productively.
Mistakes Incorrect actions committed by the The Sum of the number of errors
user while using the system, both committed by users while using the
the number of errors and their type. system or product.
Satisfaction Level The user’s opinion about the sys- The users’ assessment of their ex-
tem, considering the criteria of each perience in the use of the product.
user.
Table 1. Main attributes within Usability Engineering.
or products takes into consideration specific attributes that one can always quantify from
time records and testimony of the users who tested the operation of the product—see
Table 1. On the other hand, the usability evaluation methods are diverse, some of which
we can mention are:
– Survey: It is applicable at the initial stage of development, useful for obtaining the
requirements of the product.
– Field observation: Tests analysis of the tasks performed by the end user, in the final
stage and during the implementation of the product.
– Focus groups: This method applies to the design and requirements analysis stages.
– Heuristic evaluation. It is carried out by experts in the field, to apply at the begin-
ning of the design, during the development and before the start-up.
Towards a logic lab on the cloud: desktop and mobile sessions 57
The methods here adopted to perform the evaluation depend on the attribute to as-
sess. Besides, each method used in the evaluation, generally, involves the active partic-
ipation of the client or end user; this because it is vital to consider their opinion about
the corrections to implement in the product that they are going to use.
3.2 Security
According to [23], information security is “the protection of information and its critical
elements, including the systems and hardware that use, store, and transmit that infor-
mation.” It is also commonly referred to as the protection of confidentiality, integrity,
and availability of the components of an information system [14].
An important consideration about developing web and mobile applications is the
information security aspect. In general terms, every web application requires, at least,
to provide integrity and availability to the information it manages. In some cases, con-
fidentiality is also a desirable requirement. Providing security to applications in such
an environment is not an easy task, especially for using a public network, such as the
Internet.
3.3 Logic Programming and Answer Sets
One of the main foundations of this proposal is Answer Set Programming, ASP, charac-
terized in several kinds of non-classical logic, with a long background and suitability to
represent non-monotonic knowledge. Its main applications in problem solutions range
from typical AI toy examples to agent prototypes and planning settings. Another name
to identify this semantics from the literature is Stable Model Semantics or simply SM
for its name in the original paper [11]. In addition to that, [16,18] introduced the use of
ASP solvers as a new programming paradigm, though.
The following introduces a general description of ASP, which receives other names
as well like Stable Logic Programming or Stable Model Semantics [11] and A-Prolog.
In this paper, we assume the reader is familiar with the rest of its syntax and semantics,
easily accessible in the literature. We introduce some basic syntax, though, to get some
context of the apps.
Its formal language and some more notation are introduced from the literature as
follows.
Definition 1 (ASP Language of logic programs). In the following LASP is a formal
language of propositional logic with propositional symbols: a0 , a1 , . . .; connectives: “,”
(conjunction) and meta-connective “;”; disjunction, denoted as “|”; “←” (implication,
also denoted as “→”); propositional constants “⊥” (falsum); “>” (verum); “not” (de-
fault negation or weak negation, also denoted with the symbol “¬”); “∼” (strong nega-
tion, equally denoted as “−”); auxiliary symbols: “(”, “)” (parentheses). The propo-
sitional symbols are also called atoms or atomic propositions. A literal is an atom or
a strong-negated atom. A rule ρ is an ordered pair Head ← Body, where Head is a set
of literals and Body a set of literals and default-negated literals. Either of the two sets
may be empty.
58 D. Rodriguez-Aguilar et al.
Strong negation “∼”—also called explicit negation, constructible falsity and classi-
cal negation—in logic programs has the following meaning with respect to the default
negation “not”: a rule ρ0 ← not ρ1 allows to derive ρ0 when there is no evidence of ρ1 ;
while a rule like ρ0 ← ∼ρ1 derives ρ0 only when there is an evidence for ∼ρ1 , i.e., when
one can prove that ρ1 is false.
With the notation introduced in Definition 1, one may construct clauses of the fol-
lowing general form, well known in the literature.
Definition 2 (Extended Disjunctive Logic Program, EDLP). An extended disjunctive
logic program is a set of rules of form
`1 ∨ `2 ∨ . . . ∨ `l ← `l+1 , . . . , `m , not `m+1 , . . . , not `n , (1)
where `i is a literal and 0 ≤ l ≤ m ≤ n.
Naturally, an extended logic program (or ELP hereafter) is a finite set of rules of
form (1) with l = 1; while an integrity constraint (also known in the literature as strong
constraint) is a rule of form (1) with l = 0; while a fact is a rule of the same form
with l = m = n. In particular, for a literal `, the complementary literal is ∼` and vice
versa; for a set M of literals, ∼M = {∼` | ` ∈ M}, and LitM denotes the set M ∪ ∼M;
Additionally, given a set of literals M ⊆ A, the complement set M = A \ M.
The well-known semantics of an EDLP consists of reducing general rules to rules
without default negation “¬” because the latter get interpreted in classical logic by
using the well-known Herbrand models. In particular, the reduced rules with no default
negation Mon of a rule of the form (1) is
`1 ∨ `2 ∨ . . . ∨ `l ← `l+1 , . . . , `m , (2)
where `i are literals and 0 ≤ l ≤ m. Note that a literal is either an atom or a strong-
negated atom. The latter becomes a new distinguished literal as reduced rules towards
the Herbrand models, where complementary literals cannot be elements of the same
Herbrand model.
This kind of rules is the monotonic counterpart or positive program, also known in
the literature. Additionally, the monotonic counterpart of a set of rules is the set of the
monotonic counterparts of its rules.
Intuitively, the monotonic counterpart is where ASP can coincide with classical
propositional logic. On the other hand, default negation is precisely the main difference
between the two systems—alternately with Prolog too. As a result, the corresponding
derivation symbols, “←” for ASP and “⊃” for Classical Logic, cannot have the same
meaning.
On the other hand, all stable models are the minimal Herbrand models of a set of
first-order sentences, but not the converse. Additionally, S is a consistent answer set of
a given program P if it does not contain a complementary pair of literals.
Although we have introduced ASP as propositional (ground) programs, fixed non-
ground ASP-programs of arbitrary arity are also common in the literature. Accordingly,
non-ground ASP-programs with variables or constants as arguments are also simpler
expressions of more extended ground (propositional) ones without variables, where
Towards a logic lab on the cloud: desktop and mobile sessions 59
each ground program P is a set of its ground rules ρ ∈ P. Besides, a ground rule is
the set obtained by all possible substitutions of variables in ρ by constants occurring in
P.
Last, one may not conclude this section without mentioning that two significant ad-
vantages of ASP over other approaches. ASP has been hard work in research both on its
declarative programming framework, and at least on three efficient, competitive prov-
ing solvers with broad backgrounds: SMODELS [19], Clingo [21] and DLV [5,15,2],
which are available to run online3 . That means they can run on their server and thus,
there is no need to download and locally install the binaries or sources.
This section introduced some basic concepts needed to underpin our proposal. They
are a set of principles to satisfy by our end-product applications, and they included
methods on how to assess such characteristics, as well as a very general ASP back-
ground.
4 Design
Our proposal consists of the design, implementation, and development of a web appli-
cation and a mobile app for Android devices. Both products should exhibit the ability
to function as an online tool for researchers and interested users of logic program-
ming. With their implementation, we put forward a complete suit that meets specific
non-functional requirements, such as: Striking and dynamic graphical user interface;
intuitive operation; adaptable and scalable content; security mechanisms that do not in-
terfere with usability The web application is a complete regular browser solution, and
the mobile app is an exclusive solution for basic tasks, in circumstances where the user
needs to calculate a quick deduction on the server side, with the potential to other future
helpful features a desktop application lacks. Namely, the mobile app is an open door
for new and more robust applications, where not only does it work for small knowledge
bases and immediate straightforward deductions, but also as a massive data collection
interface suitable for knowledge representation and reasoning. Such data collections
would range from temperature, movement, pulse, geographical location, speed, to other
more complex data, such as habits, preferences, and so forth. The development of a
mobile application and a web application with consistency between the content and the
way users interact with it may seem unnecessary. However, its development shall intro-
duce a new, more practical alternative for future users and research and development,
as above suggested. Therefore, our development focuses on issues such as portability,
usability, exclusivity, and security; current modern features that ease software operation
and get to generate comfort to users.
5 Implementation
The primary operation of both applications consists of the use of a command-line text-
mode solver. As a result, it is essential to model the way in which the user can interact
3
Their respective online running versions are available at:
http://logic-lab.sourceforge.net/ which are (graphical) front-end web interfaces to
the originally implemented engines mentioned above.
60 D. Rodriguez-Aguilar et al.
with each of them. For example, if a user needs to use the DLV solver, they must first
communicate with the interface, then with the application, which requests the solver.
The processing involves a link with the communication line that links each of the parts
that make up the proposed product. In addition to that, modeling means that each solver
works as a black box, which accepts an input and returns an output. Each procedure
that comprises a functionality in the project means requests to the service. That is why,
for this flow of requests and responses, we need a controller that acts as a mediator
between the view and the model. Then, it is from this modeling that the idea of using the
controller view model (CVM) scheme to handle the data arises in detail. This scheme
organizes the process of coding or construction of the product. Besides, it facilitates
the detection of errors in the flow of requests and responses by the server or model.
This modeling, as its name suggests, it is divided into three main parts, which we can
describe below:
– Controller: it works as an intermediary between the user (client) and the server. It
controls the flow of data and transmitted requests in the communication line. That
means that if the user requests to use a particular solver, the controller makes a di-
rect request to the model defined for the handling of such ASP solver. Additionally,
it sends the necessary data for processing in the selected solver.
– Model: contains the execution lines of the inference engines and the database. In
the case of the latter, we can say that, with the help of the model, the controller
can make a CRUD request (create, update, delete), with which it can create a new
user, update data from an active one or delete a user registered in the system. For
the development of the applications that make up this project, we introduce two
models, one that controls the sessions and the other that controls the ASP solvers.
– View: It returns the result of the requests made to the model. That is to say; first,
the controller sends a request to the model, and the model returns a response, im-
mediately the controller updates the view, and the user can view it.
5.1 The Back-End
The documentation of DLV and CLINGO provides valuable information towards a
front-end and reveals that to process logic programs, it is necessary to know how to
use a command line terminal at least, for Linux/UNIX in our case study. For exam-
ple, suppose the following simple ASP knowledge base4 , {a. b D not ∼c.}. To get
the corresponding inference, one needs to send such program to the required solver
through a pipeline. A pipe is a means that accepts an entry as input and sends it out to
the subsequent command, in the pipeline, and which in turn is delimited by the use of
the character pipe “|.” Then, the process of the logic program {a. b D not ∼c.} are
expressed as follows:
echo "a. b:- not ~c." | ./ dlv.bin -- 2 >&1 | awk ’{ print $0}’
4
For the sake of economy in the pipeline command, we introduce an elementary ASP program,
which does not mean the method cannot accept thousands of program lines.
Towards a logic lab on the cloud: desktop and mobile sessions 61
where the command echo represents the escape of an output, which in this case refers
to the logic program in text mode. Such output goes through the pipeline to the DLV
solver, represented by the binary file called dlv.bin, in this case. The solver takes
the output of the previous command echo in the pipeline as input and produces a new
output, which is the derivation of the logic program. In turn, awk takes such output as
input and gives it format if required. In this case, $0 does not affect.
The description of this command line is an integral part of our proposal because it is
the central connexion to the backend, and programming languages such as PHP allow
the execution of this type of commands through the front end. This type of front end is
possible, in PHP, through the use of the shell_exec function. We describe its general use
as follows:
string shell_exec ( string $cmd)
The function accepts commands in string format, that is to say as a string of charac-
ters. Therefore, the implementation of an inference engine in a graphical environment
only requires the use of this function. It is worth noting that the execution of the logic
sentence is in a slightly different syntax, without changing its semantics.
shell_exec ("echo ’a. b:- not ~c.’ | ./ dlv.bin -- 2 >&1 ’’ | awk ’{ print $0}’")
Then, we can conclude that, by employing the previous command, it is easy to
implement inference engines such as DLV or CLINGO in a graphical environment—
web or mobile—the first phase of an online logic lab. Therefore, the primary operation
of the expected product is covered using this type of implementation.
5.2 Connections with the server and the database
A fundamental part of this development is the connection of the web and mobile ap-
plication, which gets done through the implementation of the client-server architecture.
In the first place, it is essential to specify that the client’s function gets fulfilled by the
application (web or mobile), this is because it is the one who makes the requests and
is waiting for the response. On the other hand, the server is responsible for offering ac-
commodation to the PHP script or order file, in order to provide or satisfy the customer’s
request.
The most relevant requests are, in short, the use of a specific inference engine and
the connection to the database in order to execute a CRUD action (Create, Update,
Delete) with the information of both registered users and of those about to get enrolled.
Previously, this paper introduces how ASP systems are implemented and work in appli-
cations. Now, it is necessary to explain how the connection with the database works.
The connection to the database uses the server as the host of the script responsible
for the link to the database. Next, one can see this script or snippet of code needed to
connect both applications to the database. It is possible to observe that such a connection
requires the specification of the server’s IP, the type of connection and the credentials
of access to the database (user and password).
62 D. Rodriguez-Aguilar et al.
s e t A t t r i b u t e (PDO : : ATTR_ERRMODE, PDO : : ERRMODE_EXCEPTION) ;
$ b a s e −>e x e c ( " SET CHARACTER SET u t f 8 " ) ;
r e t u r n $base ; }
c a t c h ( E x c e p t i o n $e ) { d i e ( " E r r o r : " . $e −>g e t M e s s a g e ( ) ) ; }}? >
In summary, the implementation of a database allows not only the administration of
users but also the synchronization of data between both applications. One should note
that the fact that an automatic update simultaneously occurs is an important feature
currently in the development of applications.
5.3 LogicSite Laboratory Front-End: web and mobile application
We mainly base the functionality of the applications on the integration of ASP systems.
We devote particular interest to the web application, that is why we propose an intuitive
interface. We have ordered the set of elements that are available to the user so that they
are at their hand. As we can see, the improvements within the interface, in comparison
with other existing projects, is notorious, starting with the options bar. On such options
bar, we can locate the tools of primary use in the execution of the inference engines.
On top of that, the web application integrated new features that provide greater con-
trol and accessibility for the user, understanding accessibility as the ease of performing
some action, as shown above. Among the most notable features are:
– Syntax auto-completion in reserved words: for the development of this part we have
employed an API5 called bootstrap-suggest6 , a plug-in that allows suggestions to
be displayed as the user writes inside a plain-text area. This add-on allows one
to enter all those reserved words, typical of the DLV and CLINGO systems, with
which such API searches within one’s knowledge base a match to what the user is
writing. Its operation starts from the detection of the character “#”; once the system
locates this element, it displays a list of possible options for the auto-completion of
the text in question.
– Detection of the error line: it is common that during the process of writing a logic
program, users commit syntax errors, whether a particular character is missing or a
misspelled specific reserved word. The error detection works using the output pro-
vided by the execution of each solver; so that through the result of the processing,
the solver generates the error number and the code line in which it shows up; it is
immediately highlighted, making the user’s work agiler.
5
API (Application Programming Interface) is a set of rules (code) and specifications that facil-
itate human interaction-software.
6
The bootstrap-suggest API is available through:
https://github.com/lodev09/bootstrap-suggest.
Towards a logic lab on the cloud: desktop and mobile sessions 63
– Self-saving: we include a saving function for registered users. This action is carried
out automatically as the user writes on the text area, in this way their progress gets
stored in a database identified by a unique registration number for each user. There-
fore, at each login, registered users can see their latest ASP activity they were work-
ing on, making it possible to continue with the writing of logic sentences of their
preferred solver. Once the system starts up, it updates the values of the database
contents as the user writes, and during the login session, it queries what there is in
the database.
Other noteworthy aspects of the web application are navigation within the webpage.
Users can always choose to download and use the IDEs of the ASP solvers, download
the mobile application, or enter as a registered user and send their ratings and comments
about their experience using the system.
Finally, let us introduce the development of the mobile application having the func-
tionalities described below:
– It offers the use of DLV and CLINGO inference engines in a graphical interface.
– It is possible to register and validate users.
– The password reset function we implemented takes into account that the user can
forget the password with which they got registered and, therefore, lose access to
their account. Therefore, this feature supports the user in the process of recovering
their account.
– The import and export functions of files in plain text are a fact. This functionality
allows working with external files so that the user can upload ready-made programs,
or use a text editor outside the application and load the modifications made for
future processing.
– Detection of syntax errors (by pointing out the error line), by writing suggestions
on the reserved words registered out of the knowledge base.
– Reset functions, to clean the contents of the work area and start from scratch, as
well as an auto-save of the current status of the work of registered users.
Based on this list of functionalities, it is possible to notice that options not considered
before are now part of the system and that today they are an innovation. The implemen-
tation of error detection, session management, and password recovery offer a beginning
for future innovations not only for an online logic lab but other services and applications
too.
5.4 Application of Usability Engineering Techniques
The integration of usability metrics in this project is one of the most important factors
considered within the objectives. That is why, when developing the applications, we
started establishing which usability metrics to implement. Therefore, to decide the dis-
tribution of each of its parts, the adequate colors, the symbology, and the content, the
following Usability Engineering metrics were taken into consideration:
Location context is applied to control small viewing spaces and prevent the user from
getting lost easily. For example, in the applications, we examined the information,
and the display extent we counted on, to determine the correct order and the way in
which it should get visualized.
64 D. Rodriguez-Aguilar et al.
Less is more this technique is used to control that the information contained is useful
and discard what is unnecessary. Within the applications, we avoided the saturation
of information in order not to generate confusion to the user and even to make sure
that the intuitive design prevails.
Cognitive load and visualization metric used to control the cognitive load in users,
that is, users should not memorize a large number of steps during their interaction
with applications. We designed the applications taking into consideration that the
execution and selection of the solver and the import or export of files, need to be
direct without requiring an extra effort to the user.
Block design structure plays a role in the distribution and presentation of the content
to provide an organized interface. For example, in the case of the web applica-
tion, we developed an options bar that organized the content corresponding to the
execution of the selected solver and the functions for itself. In the same way, we
examined the rest of the graphical interface for both applications to structure all the
content and improve the visualization.
Precise color and symbology implemented for the use of universal symbols associ-
ated with the action or information that one wants to convey to the user. In the
same way, the colors selected for the interface must be precise and not affect the
visualization within the applications. We applied this metric universally, that is, in
both applications, we kept the same symbology to avoid confusion among the users,
and the same happened with the designation of colors in the interface.
The implementation of these usability techniques is not exhaustive. When including
them, there is an innovation concerning the user’s experience with both applications.
Therefore, these first steps refer to specific milestones of computer-human interaction,
especially on ASP terminal-mode solvers.
Throughout the development process of this proposal, it has considered particular
features or functions implemented in other web applications. Also, new features were
added to improve the user experience within the laboratory. Thus, each of these aspects
was analyzed initially, with the intention of identifying its strengths and weaknesses,
implementing then those characteristics or functions that would strengthen this project.
5.5 Security Mechanisms
Since individual data used in the proposed applications are indeed sensitive, and be-
cause of the in-security of the means of communication used by both web and mobile
applications, it is imperative to implement security mechanisms that protect such data
from intrusions or alterations. It is also important to note that cloud and web storage
platforms depend on third parties, so it becomes nec-essary to provide security to the
applications that run on those environments. Some of the data that are considered sensi-
tive in the proposed applications, and prone to protect, are the programming codes gen-
erated by users. The access passwords to the applications and the data obtained through
mobile devices are to get protected as well. Some examples are biomedical parameters
of a patient and any other data that may acquire in the future as input to a module made
on the platform. The following describes the security mechanisms proposed for web
and mobile applications:
Towards a logic lab on the cloud: desktop and mobile sessions 65
User/Password access control. The username and password used to login on the ap-
plication are not only to control who can or cannot access the application and its
resources. It also allows users to retrieve their work previously developed and saved
in the platform. The access control also prevents intruders, or even legitimate users,
from accessing the work of another legitimate user. A username and password shall
provide a correct balance between security and usability.
Password encryption. Once a new user performs the registration process or login, the
password shall get encrypted (ciphered) using a one-way or hash function—in this
case, the named SHA-1—and then the system shall store it into the database. Such
encryption shall prevent any person, even database administrators, from having ac-
cess to the password in clear-text mode. In the login process, once the user types
in their password, the same function used in the signing process shall calculate the
hash value and compares it to the stored value for that user.
Sensitive data encryption. As previously mentioned, there are sensitive data that the
platform should protect from intrusions. Because encrypting all the information
generated and acquired by the applications reduces the computing performance,
only the data that the user considers to be sensible shall get encrypted. This prin-
ciple also applies to the ASP programs that the user writes on the plat-form. The
data shall get encrypted at the database level; that is when they are going to store
the ASP programs. This task primarily protects the confidentiality of the data.
The digital signature of ASP source codes. The ASP programs written by users are
the most crucial asset on the platform. The integrity of these programs is, there-
fore, a priority in the proposed applications. In order to protect their integrity, they
shall get a digital signature each time the user ends a session, through a public key
system. To carry out the digital signature, each user shall hold a public/private key
pair. The private key is in a container protected with the user’s password. Because
the user enters their password to log in, the system shall enable the latter to digi-
tally sign the ASP pro-gram(s) before the logout process completes. As soon as the
user’s ASP programs get stored into the database, the digital signature for each of
the programs shall get also stored. In this way, when the user starts a new session,
using their public key, the system shall perform an automatic comparison of the
stored program or programs of that user with the digitally signed versions. This
task would detect if an intruder modified an ASP program. The digital signature
process, as well as the verification of the signature, shall be transparent to the user.
The system should request the intervention of the user only if it detected that a
program does not match with the last digitally-signed version.
6 Conclusions and Future Work
The contribution of this work consists of the design of a web application and a mo-
bile app to implement the DLV and CLINGO systems, in an online friendly graphical
server-side front end. That is, taking into account the fact that, according to our research,
there is a limited number of web applications that provide such service and there is no
evidence of the existence of any ASP mobile app. In addition to that, existing apps are
limited exclusively to offering the use of such unpolished systems, without seriously
66 D. Rodriguez-Aguilar et al.
considering the user’s experience. The conclusions that we get from this research and
development work of software are the following. In the first place, we can affirm from
the implementation of the ASP logic lab system prototype, that the applications de-
veloped throughout this project allow not only to show the inferences from DLV and
CLINGO but also the manipulation of such output to become the input for future ad-
vances in the growth of the logic laboratory. Such is the case of the extraction of the
output for the detection of errors, in the input of logic sentences, which is a fact in this
project. Likewise, the session system designed for both applications perform a correct
and safe validation of the users in order to offer new functionality that until now had not
been considered in the use of ASP solvers and applications. That functionality consists
of the self-preservation of the logic clauses written by the users during their visit to the
laboratory to get continued. Therefore, the system maintains the users’ work as they left
it so that they do not have to worry about having abandoned (and lost) their activity.
In the same way, this session system shall allow maintaining synchronization be-
tween the applications, that is to say, if, during a short time the user starts writing in a
logic program in the mobile app, their work goes reflected in the web application to get
continued. The design and implementation of this function give rise to future advances
for the user’s version history so that not only could they have access to their current
work but also to resume or restore logic formulas previously deleted by themselves.
Finally, we state the fact that the development of this project opens the possibility of
more future research and development. It is worth noticing that our proposal is just a
beginning for what, in a particular moment, could be a fully-fledged laboratory for users
interested in mathematical logic. That is, with the code lines of the development of this
research we introduce a new alternative for the use of ASP solvers and systems, by
integrating sessions management to control an auto-save function and synchronization
between applications. However, in a future proposal, not only can the user experience
be reconsidered but also the manipulation of the resulting inferences for novel appli-
cations, such as in sports medicine and health, the administration of services and the
management of massive or large-scale data—big data. As we can see, the manipulation
of the information collected by these systems offers considerable possibilities in differ-
ent areas (health, education, and entertainment), which are individually exploited for
a long time already, previously mentioned in this research. In short, this project cen-
tered on changing the user interface of ASP solvers and systems, acknowledging that
the task of the user should focus more on the ASP language rather than on the difficulty
of downloading, compiling, installing and running the solver or system itself.
References
1. Acosta Guadarrama, J.: Logic Lab (2016), http://logic-lab.sourceforge.net/
2. Alviano, M., Faber, W., Leone, N., Perri, S., Pfeifer, G., Terracina, G.: The Disjunctive
Datalog System DLV. In: First International Workshop, Datalog 2010, Oxford, UK, March
16–19, 2010. Revised Selected Papers. pp. 282–301 (2010). https://doi.org/10.1007/978-3-
642-24206-9_17, https://doi.org/10.1007/978-3-642-24206-9_17
3. Arieta, V.: GuessAndCheckers (2016), https://github.com/vincenzoarieta93/
GuessAndCheckers
4. Calabria, U.: DLV (1997), http://www.dlvsystem.com/dlv/
Towards a logic lab on the cloud: desktop and mobile sessions 67
5. Calimeri, F., Dell’Armi, T., Eiter, T., Faber, W., Gottlob, G., Ianni, G., Ielpa, G., Koch, C.,
Leone, N., Perri, S., Pfeifer, G., Polleres, A.: The DLV System. In: Flesca, S., Ianni, G. (eds.)
Proceedings of the 8th European Conference on Artificial Intelligence (JELIA). Springer,
Cosenza, Italy (September 2002)
6. Calimeri, F., Fuscà, D., Germano, S., Perri, S., Zangari, J.: EMBASP (2015), https://www.
mat.unical.it/calimeri/projects/embasp/
7. Calimeri, F., Fuscà, D., Germano, S., Perri, S., Zangari, J.: A Framework for Easing the De-
velopment of Applications Embedding Answer Set Programming. CoRR abs/1707.0 (2017),
http://arxiv.org/abs/1707.06959
8. Campisano, D.: DLVfit (2015), https://github.com/brainatwork/DLVfit
9. Garcia-Mata, C., Marquez, P.: Answer Set Programming y el Problema de Asignación de
Frecuencia, FAP. ELECTRO (2009)
10. Gebser, M., Kaminski, R., Kaufmann, B., Schaub, T.: Clingo = ASP + Control: Prelimi-
nary Report. Tech. rep., Aalto University, Finland; University of Potsdam, Germany (2014),
http://arxiv.org/abs/1405.3694
11. Gelfond, M., Lifschitz, V.: The Stable Model Semantics for Logic Programming. In: Kowal-
ski, R.A., Bowen, K.A. (eds.) 5th Conference on Logic Programming. vol. 88, pp. 1070–1080
(1988)
12. Germano, S., Calimeri, F., Palermiti, E.: LoIDE: a web-based IDE for Logic Programming
Preliminary Technical Report. CoRR abs/1709.0 (2017), http://arxiv.org/abs/1709.
05341
13. Germano, S., Palermiti Eliana, C.F.: LoIDE (2016), https://github.com/
DeMaCS-UNICAL/LoIDE
14. ISO/IEC: ISCO/IEC 27002: Code of practice for information security management (2005),
www.iso.org
15. Leone, N., Pfeifer, G., Faber, W., Eiter, T., Gottlob, G., Perri, S., Scarcello, F.: The DLV
system for knowledge representation and reasoning. ACM Transactions on Computational
Logic 7(3), 499–562 (2006). https://doi.org/http://doi.acm.org/10.1145/1149114.1149117
16. Marek, V.W., Truszczynski, M.: Stable models and an alternative logic programming
paradigm. In: V.W., M., M., T., D.S., W. (eds.) Artificial Intelligence, chap. The Logic,
pp. 375–398. Springer, Berlin, Heidelberg (1998). https://doi.org/10.1007/978-3-642-
60085-2_17, http://link.springer.com/10.1007/978-3-642-60085-2_17http:
//arxiv.org/abs/cs/9809032
17. Nelsen, J.: Usabilidad, diseño de sitios web. Prentice Hall, 1 edn. (2000)
18. Niemela, I.: Logic programs with stable model semantics as a constraint program-
ming paradigm. Annals of Mathematics and Artificial Intelligence 25(3/4), 241–
273 (1999). https://doi.org/10.1023/A:1018930122475, http://link.springer.com/
10.1023/A:1018930122475
19. Niemela, I., Simons, P.: Smodels—an implementation of the Stable Model and Well-
Founded Semantics for normal logic programs. In: Proceedings of the 4th LPNMR (’97).
LNCS, vol. 1265, pp. 420–429. Springer, Dagstuhl Castle, Germany (1997)
20. Peinado Portillo, M.A.: Modelado de Decisiones Sesgadas en Answer Set Programming.
In: Acosta-Guadarrama, J.C., Rodas Osollo, J., Ramírez-Bouchot, M. (eds.) RCCS. pp. 1–8.
CEUR, ISSN 1613-0073 (2016), http://ceur-ws.org/Vol-1784/99990001.pdf
21. Potassco, U.P.: Clingo (2016), https://potassco.org/clingo/
22. Potsdam University, P.: Running Clingo (2017), https://potassco.org/clingo/run/
23. Whitman, M.E., Mattord, H.J.: Principles of Information Security. Course Technology Press,
Boston, MA, United States, 3rd edn. (2007)