Degenerative diseases, such as the Alzheimer's Disease, can be very serious and life-threatening. As the scienti c community strives to fully understand their exact root causes and advance their research on the domain, a massive amount of knowledge is generated. To represent and link all this knowledge, we propose the generation of knowledge graphs from the scienti c literature. We aim to provide researchers the ability to relate their new discoveries with the current knowledge and possibly formulate new hypotheses to further advance the research. In this paper, we describe a method to extract information from scienti c literature for generating a knowledge graph reusing existing domain ontologies. We demonstrate the e ectiveness of our method by generating knowledge graphs from a set of abstracts of scienti c papers on Alzheimer's Disease.
Degenerative diseases a ect the function and structure of cells, tissues, and
organs, becoming worse over time [
One example of such neurodegenerative diseases is the Alzheimer's Disease
(AD). It is the sixth leading cause of death in the United States, and the fth
leading cause among individuals aged 65 and older [
AD is still not curable. The current treatments for the disease aim on slowing its progress down in the a ected individuals, for which an early diagnosis is of extreme importance. The scienti c community works on better understanding the disease to nd new methods of early diagnosis, treatment, and ultimately, the cure. This research on AD continuously generates new knowledge. Integrating available data and properly representing the domain knowledge could bring great bene ts. It could, for instance, provide the researchers the ability to visualize how the known concepts and their discoveries may relate to each other, as well as correlate their ndings to discoveries from other researchers. By observing such relations, researchers might be able to formulate new hypotheses, and in this way, advance the domain's current state-of-the-art.
Knowledge Graphs (KGs) de ne the interrelations of real world entities in
facts represented as a graph [
In this article, we investigate the generation of KGs as automatically as possible from the scienti c literature on AD. Several research challenges remain open in our study context. The scienti c literature has a speci c, yet not uniform writing style, then posing several issues to information extraction. There are cases where the sentences are too long, containing several abbreviations, and a very speci c set of terms that are only known by domain specialists. Natural Language Processing (NLP) tools involved in the information extraction techniques are usually not trained for applications in such vocabulary. For those reasons, a completely automatic system to generate KGs from scienti c texts is a hard task to be accomplished.
Our goal is to de ne a (semi)-automatic method to generate KGs from the processing of unstructured text obtained from scienti c papers on the AD domain. Via our method a KG is generated by the identi cation and extraction of information from unstructured text using NLP techniques. The extracted information is stored in the form of RDF triples. Then, we identify the concepts and relations present in the text using a knowledge base mapped to a single domain ontology that is recommended by the NCBO2 bioportal. Via the implementation of a KGen software tool3, we evaluate the e ectiveness of our approach to generate KGs from scienti c papers on AD.
The remaining of this paper is organized as follows: Section 2 discusses the related work; Section 3 presents our method, along with a running example to illustrate the process; Section 4 shows the evaluation with KGs generated from a
set of scienti c papers; Section 5 discusses our obtained ndings. Finally, Section 6 closes the paper presenting the conclusions and future work. 2
The generation of KGs from unstructured texts has been studied in the past
years for the purposes of knowledge representation and reasoning. The
knowledge extraction of RDF triples was addressed via the use of open information
extraction systems, such as ReVerb [
Di erently, Exner and Nugues [
Martinez-Rodrigues et al. [
Similarly, T2KG tool [
Other software tools have been proposed for the purpose of KG building. For
instance, the IBM provides a tool for the information extraction from plain text
to ultimately build a KG integrating input documents [
Concerning KGs and AD, Lam et al. [
The National Center for Biomedical Ontology [
The vast majority of the investigations dealing with RDF in the biomedical domain focus on the ontologies, either by creating, nding, and unifying them. In our case, we focus on generating a KG from a text. In this process, entities are then mapped to already existing ontologies. We seek to match instances of concepts and entities from ontologies in a text, which may even describe information that is not yet captured in an existing ontology, due to its novelty aspect. The works presenting techniques to generate KG are not applied to the biomedical or scienti c domains, and we seek to employ similar techniques to address the generation of KGs on scienti c texts from this domain. 3
We describe the proposed KGen method developed to generate KGs from
unstructured text. KGs rely on RDF and Linked Data principles [
Formally, a Knowledge Graph KG = (V; E ) is represented as a regular graph with a set of Vertices V and Edges E . Whereas the vertices express entities or concepts, the edges express the relations between them. A RDF triple refers to a data entity composed of subject, predicate and object de ned as t = (s; p; o). In KGs, the relations are predicates (p), such that E = fp0; p1; :::; png, i.e., the edges in KGs are a set of predicates. The predicates are formally de ned in ontologies.
An ontology O describes a domain in terms of concepts, attributes and
relationships [
Also in KGs, the entities or concepts are either subjects (s) or objects (o), considering that the vertices are a set of subjects and objects, such that V = fs0; s1; :::; sn; o0; o1; :::; ong. In this context, oi 2 CO. We may also say that a KG is a set of RDF triples, such that, KG = ft0; t1; :::; tng, where t0 = (s0; p0; o0); t1 = (s1; p1; o1); :::; tn = (sn; pn; on).
Figure 1 presents the key elements involved in our method. First, there is a preprocessing of the input text, in a way to identify all the sentences available. From such sentences, our method of KG generation extracts triples by identifying the subject, predicate and object. Afterwards, it performs the identi cation of entities, concepts and properties from the sentences to obtain links in our graph to an ontology. By combining the output of both steps, i.e., the triples and the ontology links, the method reaches a new set of linked triples. Finally, the generated graph is represented in RDF turtle7 format.
Preprocessor. The preprocessing step (cf. 1 in Figure 1) receives as input an unstructured text le in plain-text format (e.g., a *.txt le). This preprocessing step submits the text through some sub-steps. The rst sub-step identi es
the sentences from the raw text using a sentence splitter. This NLP tool outputs each identi ed sentence per line. Afterwards, the next sub-step resolves co-references by using a NLP technique which identi es pronoun references in di erent sentences (e.g., John lives next door. He works on Sundays. { He refers to John). Once identi ed the references, the pronouns are changed to the actual references (e.g., John lives next door. John works on Sundays ). This is important to keep the coherence when generating triples, maintaining the actual entities in the subjects and objects, instead of the pronouns.
The next preprocessing sub-step is an abbreviation resolver. The method identi es a common practice in scienti c writing: abbreviations of terms when they are rst mentioned in the text, whereas the abbreviation is used from that point on. For instance, Alzheimer's Disease (AD) can be very serious and lifethreatening. AD is the sixth leading cause of death in the United States. { In this case, Alzheimer's Disease is replaced in the remainder of the text by AD. The substep replaces the abbreviated form by the original one in every identi ed instance to generate coherent triples.
The last preprocessing sub-step aims at simplifying sentences. This is also a common practice in scienti c writing, where we may have a complex sentence bound by conjunctions (e.g., Mitophagy inhibits amyloid-beta and thau pathology ) { In this case, it would be preferable to have two distinct sentences (e.g., Mitophagy inhibits amyloid-beta and Mitophagy inhibits thau pathology ) to generate coherent triples. The overall output of the preprocessing step is a text containing a simpli ed, co-reference and abbreviation-resolved sentence.
Figure 2 presents a running example to illustrate our de ned process. It presents the unstructured input text and its correspondent preprocessed output.
Extractor of triples. The next step is the extractor of triples (cf. 2 in Figure 1), which takes as input the preprocessed text. In this step, each sentence is processed to identify the candidate predicate, the subject and the object. Our proposal explores a Semantic Role Labeling (SRL) technique to perform this identi cation as the rst sub-step. SRL identi es the verbs from a sentence, along with its agents, patients and other semantic roles (e.g., theme, topic, etc.).
Once agent and patient are identi ed, the second sub-step explores an
algorithm to identify the triple's constituents. According to the triple de nition,
t = (s; p; o), the method needs to identify the subject, predicate and object from
the SRL output. The predicate is naturally mapped to the verb. As for the
subject and object, if there is an agent and a patient linked to the verb, the subject
maps to the agent and the object to the patient. It is important to mention that,
in case of passive voice sentences, the patient and agent may be out of order,
but SRL already assigns them correctly. If there is an agent, but no patient, the
subject maps to the agent, and the object maps to the closest semantic role to
patient. The same applies in the case we have a patient (mapped to the object),
but no agent (subject maps to the closest semantic role). Finally, if either the
object, or the subject cannot be mapped to a role, we discard the sentence,
as, per de nition, a triple must have its three constituents. This algorithm has
been adapted from a similar algorithm de ned by Martinez-Rodrigues et al. [
Ontology linker. This step (cf. 3 in Figure 1) takes as input the preprocessed text. The rst sub-step performs a tokenization to split the sentences into tokens in addition to explore a Part of Speech (PoS) Tagger to tag the tokens (e.g., as nouns, verbs, adjectives, etc.). Then, a parse tree is obtained for the referred sentence. By doing so, the technique enables the identi cation of verbs, considered as predicate candidates, and noun phrases, considered as subject/object candidates. Such candidates are matched against an ontology to nd correspondences on such ontology's concepts and attributes.
Formalizing this process, a sentence S = ft0; t1; :::; tng is a set of terms (or tokens) ti. Each term gets a PoS p associated (ti; pi). A predicate candidate pc = tijpi = \V B00 is a term whose PoS is a verb. A subject/object candidate soc = ftig is a set of terms whose parents in the parse tree are noun-phrases (NP ). Each candidate is then associated to ontology elements, i.e., pc ) RO, and soc ) CO.
Graph builder. The graph builder (cf. 4 in Figure 1) takes as input the triples and the ontology links. In this step, the method rst creates a local resource for each of the triple's constituents, binding them to resources obtained from the ontology links. This results in a turtle content describing the KG. We convert the turtle content to a set of vertices and edges, feeding them to a graph generator system, which outputs the graph. (rdf:Property ). The subject and object types are classes (rdf:Class ). The subject is linked to an ontology resource (nci:C2866 ), whereas the object is not.
Implementation aspects. KGen has been implemented in Python
language. In the preprocessing, we used Stanford's CoreNLP [
The ontology linker uses mainly Stanford's CoreNLP [
The goal in the evaluation of our method is to understand the quality of the generated KGs. For this purpose, we used as input for the method abstracts of scienti c papers dealing with AD, obtained from PubMed 10. The KGs generated, along with their intermediary artifacts, are available in the KGen project repository11. The linked ontology that better suited the abstracts was the National
Cancer Institute Thesaurus 12 (NCIT). It was the ontology returned by NCBO's recommender endpoint13 for all the abstracts.
The subjects and objects of the triples are most of the time composed of more than a single entity (or noun-phrase). To capture such characteristic, the local partof property (local:partof ) links the composing entities belonging to a subject or object. This property, in turn, is linked to NCIT's Part Of property (nci:C43743), as illustrated in Figure 4. This sub-graph is available in all graphs generated for this evaluation, as they all have composing entities (e.g. \series" and \rutacecarpine derivatives" are composing entities of \A series of rutacecarpine derivatives").
The snippet of the KG present in Figure 5 has been generated for the speci c triple t=(A series of rutacecarpine derivatives, identify, novel ligands), obtained from one of the abstracts14. In this graph, the original triple is represented by linking the predicate to the subject and object through the rdf:subject and the rdf:object properties. This form of rei cation was chosen to allow the representation of the constituent parts (local:partOf ), and links to the ontology concepts and attributes.
The local entities and properties are represented preceded by the local pre x (e.g., local:identify ), and their original text is mapped through the rdfs:label (e.g., identify ). As literals, those values are represented inside rectangular nodes. The other nodes, as resources, are represented inside elliptical nodes (cf. Figure 6).
The links to the concepts and attributes of the ontology are achieved through the rdf:type property. They are represented by their ontology pre xes and code property (e.g., nci:C25737 ). To enhance the readability, we present their preferred names, through the rdfs:label property (e.g. Identi cation). In case no link was retrieved in the ontology for an entity, their local resource is not bound to an ontology resource in the graph, as illustrated in Figures 5 and 6. 12 https://ncit.nci.nih.gov/ncitbrowser/ 13 https://bioportal.bioontology.org/recommender 14 https://www.ncbi.nlm.nih.gov/pubmed/31136894 This investigation aimed to generate KGs from the scienti c literature on degenerative diseases. Our method linked the concepts, entities and properties from the graph to classes and attributes from existing ontology in the biomedical domain. The way of combining extracted triples with the ontology linkage, both from this speci c domain, refers to the key originality in this investigation. Our ndings indicate success in generating KGs for unstructured text from abstracts of scienti c papers.
The language employed on scienti c papers, especially those in the degenerative diseases domain, pose a great di culty for the techniques and tools involved in the method. For this reason, a fully automated method is still a challenge. Although our method is able to run to completion without human intervention, the method allows a domain specialist to review and manually change the intermediate artifacts, i.e., the preprocessed text, triples, ontology links, and the RDF representation of the KG. In the KGen tool, such intermediary artifacts are represented by text les. When they are manually changed, the tool is able to reconsider those intermediary les and update the graphs.
Some aspects of our work demand further improvements. The triples generated from the text sentences capture the most important aspects dealt within the text, but secondary information is usually left aside from them. Open Information Extraction systems and Semantic Role Labeling focus mainly in the verbal relations. Secondary information, not directly related to the main verb is, therefore, not captured in the KG. We believe that exploring the output of a dependency parser could bring into the graph such missed information.
The linked concepts and properties from an ontology requires additional improvements. We could explore alternatives to nd a link when there is not an exact match. In order to minimize the cases where no link is assigned to a local concept, we plan to investigate SPARQL queries to obtain more generic concepts within an ontology, or search for a match from another ontology and then seeking to nd a mapping between these two ontologies.
We plan investigating alternatives to improve the issues and re ne our method to generate an ontology-linked KG from scienti c documents. Domain specialists will be invited to assess the obtained KGs. 6
The creation of KGs from scienti c literature on degenerative diseases can help researchers investigating how their discoveries relate to the existing domain, and to other researchers' discoveries. However, the automatic generation of KGs is an open research challenge. In this article, we proposed a method to generate ontology-linked KGs from scienti c papers on degenerative diseases. Our method is suited to extract triples and connect them with existing ontologies. The conducted evaluation used abstracts obtained from scienti c papers. We showed that the KGs were successfully generated from them. Future work involves generating KGs linked from di erent ontologies, as well as studies comparing temporal texts through their generated KGs.
This work is supported by the S~ao Paulo Research Foundation (FAPESP) (Grant #2017/02325-5)15. 15 The opinions expressed in this work do not necessarily re ect those of the funding agencies.