How can i best distinguish between what is fact and what is interpretation/conclusion by a historian when reading a book or article?

It should be noted that all "facts" are, to some degree, interpretations. This is beyond the very obvious point that the choice of which "facts" to include is part of the interpretative work of the historian. For example:

On August 6, 1945, the US dropped an atomic bomb on Hiroshima, Japan.
On August 9, 1945, the US dropped an atomic bomb on Nagasaki, Japan.
On August 15, 1945, the Emperor of Japan announced his acceptance of unconditional surrender.

...consists of "facts" in a sequence, and looks pretty straightforward. Thus it is easy to infer or conclude that the atomic bombs ended the war. But what have I not included? A near infinite amount of things. If I rely on a few different lines, none less factual than before, the narrative changes dramatically:

In the summer of 1945, elements in the Japanese high command, with the sanction of the Emperor, attempted to open up a communications channel with the Soviet Union, with the hope that they might negotiate a mediated peace with the United States.
But on August 8, 1945, the Soviet Union declared war on Japan and invaded Manchuria.
On August 15, 1945, the Emperor of Japan announced his acceptance of unconditional surrender.

Which leads to a very different conclusion about what ended the war. Which narrative is more "factual"? Well, neither. They're both equally factual. But historians will argue about which facts are the most useful for understanding the end of the war. They are necessarily a subset of the total number of "facts" that could be presented. My point: the historical narrative itself is always a form of interpretation. Different historical arguments are made by marshaling different sequences of "facts." And there isn't any way out of this: you can't have a narrative that contains all possible facts, for a lot of reasons (a. it would be terrible and boring, b. we don't actually have access to literally all facts/evidence, only a subset, c. many of those facts would be irrelevant and/or misleading — arguments aren't there just for show, they're there to help us understand the past, and irrelevant/misleading facts don't help with that, even if they are facts).

So that's one thing. The other is that what is and is not a "fact" is itself an interpretive action. Sometimes that interpretive leap is one pretty much everyone is willing to take: how do I know that the Soviet Union declared war on Japan and invaded Manchuria? Well, there are records, lots of them. It was reported in newspapers. There are records from the Russian archives. There are recollections from the Japanese who took part in it. There is a paper trail of intercepted communications between the Japanese and the Soviets. There are graves of people who died in the Manchurian offensive. And so on. So nobody is really going to contest my interpretation of these records to say, "that happened." It is interpretive, but uncontroversial.

But there are many cases in which the evidence is much more controversial. The example I always go to is the so-called Hiroshima warning leaflets. These are leaflets that appear to warn the Japanese about atomic bombs. They show up on many websites. They are legitimate documents in that they are not forgeries or anything like that. But people infer the wrong information about them very readily: they incorrectly assume that because such leaflets exist, they were dropped on Hiroshima and/or Nagasaki as a means of warning the citizens to evacuate before the bombs went off. They were not. In fact, the document itself cannot tell you that kind of information — you need more evidence to tell you what the evidence itself is. Historians call this "contextualization": the evidence that tells you how to understand the evidence. (And obviously this can be a regressive loop endlessly — the evidence that tells you about the evidence that tells you about the evidence...) So in this case, there is a report that describes how these leaflets were created and used — and how they were not used. Specifically, none of these leaflets were dropped on Japanese cities before both Hiroshima and Nagasaki were bombed. The leaflet itself is a post-Hiroshima creation, and Nagasaki didn't get any leaflets until the day after it was bombed. (Read the link if you want the full back-story, I am just using this as an example.)

Which is just to say: the distinction between "facts" and "interpretations" is a pretty slippery one if you are a professional. All "facts" are interpretations to some degree. The question is, how large of a "leap" is the interpretation? Some are big leaps, some are small. Can you spot a "big" versus a "small" leap? Sometimes: words like "probably" and "maybe" and "perhaps" are often used to subtly indicate where the author is inferring things he or she does not or can not know. In some cases, you need to be a subject-matter expert to pick these things apart — there is just not going to be any way that you can read a history of the atomic bomb, for example, and see the same things that I do, where I can more or less instantly tell if something is a reliable interpretation or a big jump, or where something is a new idea or an old one, or what is based on a document and what is extrapolated. I can only do this because I have been studying this topic for a decade. The best you can do is either to read a lot of other things so you, in some miniature way, replicate that sort of expertise (which is to say, read what I have read, become an expert yourself), or you can read me (or someone like me) commenting on these things and explaining to you how an expert would read it (which is the function of book reviews and the like).

I apologize if that sounds "elitist" but such is the nature of expertise. Historical knowledge is written in narratives and thus can appear more "accessible" than, say, a paper on theoretical physics or medicinal trials. But it is still a form of expertise, and there are no easy "short cuts" to becoming an expert.